MICRO

MATTERS Using Data for Development in the

Era of the Fourth Industrial Revolution

T
INDIA 2023
We thank the following organisations and their teams for sharing their stories and insights
 MICRO
MATTERS
Using Data for Development in the
Era of the Fourth Industrial Revolution
© 2023 Reliance Foundation and Observer Research Foundation.
All rights reserved. No part of this publication may be reproduced 
or transmitted in any form or by any means without permission
in writing from Reliance Foundation and ORF.

Attribution: Jayashree B, Anirban Sarma, Vanita Sharma,

and Shoba Suri, Eds. M
 icro Matters: Using Data for Development in the
Era of the Fourth Industrial Revolution, March 2023, Reliance Foundation
and Observer Research Foundation.

ISBN Digital: 978-93-90494-77-4

ISBN Print: 978-93-90494-76-7
 MICRO
MATTERS
Using Data for Development in the
Era of the Fourth Industrial Revolution
Editors
Jayashree B
Anirban Sarma
Vanita Sharma
Shoba Suri

Authors
Antara Sengupta
Ketaki Hate
Mohini Ganguly
Noyontara Gupta
Saji Kadavil
Shimona Mohan
Sitara Srinivas
Swati Prabhu
CONTENTS

I. INTRODUCTION 9

II. CASE STUDIES

Data-Informed Decisions: The ‘ASMAN’ Intervention for Better Maternal 16
and Newborn Care
by Ketaki Hate

‘CottonAce’: Weaving AI into Agriculture 24

by Shimona Mohan

E-Solutions as a Way to Empower Communities with Data 32

by Sitara Srinivas

‘Kisan Diary Enterprises’: A Mobile Solution for India’s Small Farm Sector 40
by Saji Kadavil

Enabling Disaster Preparedness with Geospatial Data 48

by Mohini Ganguly

‘Mindspark’: Working to Deliver Edtech for All 56

by Antara Sengupta

The ‘Path’ to Ushering a New Era for Digital Health 66

by Noyontara Gupta

‘Cloth for Work’: Using Data to Build a Circular Economy 74

by Swati Prabhu

III. OUR APPROACH 81

IV. ACKNOWLEDGMENTS 82
INTRODUCTION

In November 2022, on the

assumption of the G20 presidency, Prime Minister
eve of India’s must be harnessed creatively to improve public
service delivery. Subsequent presidencies have
Narendra Modi announced that the principle of all asserted that the smart use of the growing
‘data for development’ (D4D) would be integral wealth of data produced by digitalisation could
to India’s tenure. His statement was echoed by
1 greatly strengthen decision-making and policy
the G20 Bali Leaders’ Declaration, in which heads outcomes. A corollary of this—also consistently
of state unanimously “reaffirm[ed] the role of data emphasised by G20 nations—has been the need
for development” in promoting “economic growth to boost technologies for data collection and
and social well-being”.2 analysis. In this regard, artificial intelligence (AI),
big data analytics, the Internet of Things, and
The importance of D4D, and more broadly of other emerging technologies continue to push the
fostering a data-driven culture, has since been frontiers of the possible.
foregrounded repeatedly by leading state
representatives—by India’s G20 Sherpa, who has India’s status as a global digital powerhouse makes
spoken forcefully of how data-led governance it an important voice in the evolving discourse on
helped transform the country’s aspirational districts; D4D. Indeed, data-driven applications are intrinsic
and also by a host of Indian business and civil to the Digital India programme, which is building
society leaders. While the use of D4D is not a novel
3 a “digitally empowered society and knowledge
approach per se, stakeholders have been quick economy”, and recalibrating “governance and services
to grasp that new technologies allow data to be on demand”.4
gathered, processed, and applied in ways that were
not possible earlier. It was fitting, therefore, that The 2022–23 period—the run-up to India’s G20
the very first side event of the G20’s Development presidency—witnessed the launch of several
Working Group under the Indian presidency potentially game-changing data initiatives. In May
sought to explore how a tech-based D4D 2022, for instance, the NITI Aayog unveiled the
orientation might advance efforts to achieve the National Data and Analytics Platform (NDAP) and
2030 Agenda. Clearly, a new wave of interest in opened it to the public in order to improve access
the power of data has been unleashed, and a new to and the use of government data.5 The NDAP is
moment is at hand. a user-friendly web platform that aggregates and
hosts datasets from across India’s vast statistical
The zeitgeist has been several years in the making. infrastructure. It is democratising data delivery by
Since the Japanese G20 presidency of 2019, making government datasets readily accessible,
there has been a growing recognition that data implementing rigorous data-sharing standards,

9
enabling interoperability across the Indian data generation has led the World Economic Forum
landscape, and providing users with helpful tools to estimate that drones and the data economy
for analysis and presentation.6 they are establishing could boost India’s GDP by
US$100 billion and create nearly half a million jobs in
In the same month, the Indian government coming years.11 And the United Nations has
published a draft National Data Governance recognised India’s data-based SDG India Index
Framework Policy (NDGFP) that aims to ensure that and Dashboard as a “crucial tool in India’s SDG
non-personal and anonymised data from government monitoring efforts” and an “advocacy tool
and private entities can be made available to the to propagate the messages of sustainability,
country’s innovation ecosystem. Once the NDGFP 7
resilience and partnerships”. 12

is finalised, the government intends to share

anonymised datasets collected and harmonised Finally, India is rapidly emerging as a leader in the
under the framework with India’s research, startup use of geospatial technologies, or the tech-led
and AI communities. The database could be used
8
acquisition of data referenced to the Earth and its
by companies to train their AI models, and it is use for analysis, modelling, and visualisation.
expected that access to this rich and varied corpus Addressing the UN World Geospatial Information
of data will spur innovation, have far-reaching Congress in October 2022, Prime Minister Modi
impacts on development projects at all levels, and declared that “geospatial technology has been
usher in a new era of digital governance. driving inclusion and progress” and went on to
outline how it had become a key enabler of
Undoubtedly, one of India’s most eagerly development.13 A month later, India put in place
anticipated data-related developments of 2022 its far-sighted National Geospatial Policy, with a
was the introduction of the Digital Personal Data set of ambitious milestones to be met by 2035.14
Protection Bill in November. Recognising the
importance of data for the development of the The tactical use of “micro data” or sub-state-level
digital economy, the new Bill strikes a sensitive data—whether related to individuals, households,
balance between the rights of individuals and the villages, cities, or districts—lies at the heart of
interests of businesses, which may lawfully use the bottom-up development approaches. This
personal data of individuals. While the Bill does gives philanthropic institutions and civil society
not foreclose the use of personal data, it stipulates organisations (CSOs) a natural role as a bridge
that data can be processed only after organisations between communities and local governance
have sought consent from individuals, or in cases initiatives on the one hand, and the government’s
where such consent can be assumed. Several other macro vision and steps to promote D4D on the
clauses extend substantial rights to individuals other. Indeed, with their on-ground presence,
and offer them more decisional autonomy and entrenched relationships with community-
control over their data than ever before.9 level stakeholders, nuanced understanding of
local contexts, and management heft, CSOs are
India’s advancements in the domain of data use uniquely placed to execute D4D projects in
and its commitment to D4D are being lauded on partnership with administrative bodies, or to
the world stage. The 2022 BRICS Forum on Big achieve outcomes that support the latter’s work.
Data for Sustainable Development saw the five
member states agree to enhance collaboration around What often makes CSOs’ D4D interventions
big data as a tool for sustainable development, and extraordinary and extraordinarily challenging is
Indian voices were instrumental in helping the Forum the velocity of technological change they have
agree that increased data-centric cooperation contend with. As Klaus Schwab points out in The
among nations is imperative. The country’s prowess
10
Fourth Industrial Revolution, “Contrary to the
at deploying drones for development-focused data previous industrial revolutions, this one is evolving

10
at an exponential rather than linear pace. This is 2. Communicate real-time data and trends to
the result of the […] fact that technology begets decision-makers through dashboards
newer and ever more capable technology.” Schwab
goes on to observe that the speed and scope of Even as data-based insights shape the field-level
change apart, “the fourth industrial revolution is execution of projects, real-time data analytics, when
unique because of the growing harmonization made available to decision-makers, enable policy
and integration of so many different disciplines decisions and course correction. Dashboards and
and discoveries.”16 This then, is the milieu in which other e-platforms are increasingly being used to
CSOs—and every other development actor—must track and display data, identify trends, and decode
strive to make a difference. volatility with impressive results. For instance, under
the Alliance for Saving Mothers and Newborns
Micro Matters explores eight interventions by CSOs (ASMAN) project in Rajasthan and Madhya
in India that are advancing the country’s D4D Pradesh, officials are using a dashboard to monitor
agenda by gathering data; processing it to performance indicators related to maternal and
evolve insights; translating insights into actions; neonatal care at health centres. This helps them
and making a social impact. Their modus operandi take swift remedial action when needed, and is
reflect many of the priorities articulated by the improving the quality of care for mothers and
government, and are examples of the D4D ethos that newborns. Going forward, the construction of
India is keen to promote as president of the G20. dashboards ought to become a core element of
D4D projects, given their impact on transparency,
As the eight case studies in this volume traverse the accountability, and efficiency.
expanse of India, eight lessons emerge:
3. Optimise the use of emerging technologies
1. Amalgamate tech-based data collection with the
use of legacy datasets and knowledge Emerging and disruptive technologies hold
great promise for national development, and
To offer actionable insights, the data being geospatial technologies and AI are chief among
collected needs to be focused but wide-ranging, them. The country instituted its National Geospatial
offering a multifaceted perspective of the issue Policy in December 2022; the draft National
under consideration. While tech-based methods Strategy for Artificial Intelligence explores
are often used to gather data, a second layer the application of AI to healthcare, education,
of data generation—drawing on secondary agriculture, smart cities, and transportation;17 and
sources, legacy datasets, and even focus group the Union Budget 2023 included a special focus
discussions or stakeholder interviews—is crucial. on how to “make AI work for India”.18 A new
Digital Green’s Kisan Diary Enterprises (KDE) generation of Indian D4D initiatives is beginning
mobile application, for example, is used to to leverage emerging tech with powerful
collect primary details about members of farmer outcomes. As part of its disaster preparedness
producer organisations (FPOs) and the crops work in Assam’s flood-prone Cachar district,
they cultivate. But it is only when this information Reliance Foundation has geo tagged the
is collated against a large body of data about district’s critical infrastructure and public
FPO profiles, harvest patterns, cultivation facilities where community members can be
practices and market trends gleaned from state sheltered or evacuated during natural disasters.
agriculture departments, livelihood missions And Wadhwani AI’s app for cotton farmers,
and frontline workers, that KDE’s strategies for CottonAce, uses AI to analyse images of cotton
boosting FPO productivity can effectively be pests and other details uploaded by the farmers
implemented at scale. and generates instant agricultural advisories
for them.

11
4. Incorporate ancillary services into the design of (DDMA) retains all personal survey data and has
D4D interventions expressed its intention to share only anonymised
data with trusted partners. And Wadhwani AI tries
D4D programmes afford opportunities to to minimise the amount of personal information
strengthen ecosystems by rolling out targeted required by its CottonAce app, and pushes for
ancillary services. These services, such as the data compliance through stringent non-disclosure
systematic sensitisation and capacity-building agreements with partners.
of stakeholders, should be built into the very
design of the intervention. For instance, Anudip 6. Visualise scalability and replicability from
Foundation’s data-led efforts to empower women conception
digitally became an occasion for the organisation
to sensitise its partners and field workers about Tech-enabled D4D projects have an inherent
data hygiene and cyber fraud. The Mindspark capability to be scaled and reproduced. It is prudent,
edtech platform enhances learning outcomes for therefore, to visualise outcomes at scale from
schoolchildren using data, but its creators also the outset. Several cases bear this out. PATH’s
run associated teacher training programmes and public-private agency programme began as a
cybersecurity awareness campaigns for parents. data-centric initiative to diagnose and treat
Similarly, the e-training tools built into the tuberculosis (TB) patients in Mumbai. Its digital
ASMAN app have encouraged self-paced learning approach allowed the government to scale its model
among healthcare providers, boosting their nationwide. Today, PATH’s learnings from the local
confidence and allowing them more time for and national levels are helping it strengthen the
patient care. pan-India TB response, refine a private sector
engagement model that can be applied across
5. Build trust by ensuring data confidentiality health challenges, and operate a pilot with the
and security National Health Agency to bring private healthcare
providers on board the country’s Digital Health
The credibility and success of D4D projects Mission. In a similar manner, the innate scalability
will hinge on how well they maintain data and inspiring results of the Mindspark programme
confidentiality and security, and ensure an that began in Lucknow have prompted the NITI
individual’s fundamental right to privacy. Several
19
Aayog to replicate it in other parts of Uttar
organisations interviewed for Micro Matters Pradesh.
observed that there is a strong perception that
the government is the most dependable custodian 7. Offer alternate solutions where connectivity and
of data. As a result, arrangements whereby data is access are a challenge
stored by government entities tend to generate a
higher degree of confidence among citizens. This Poor internet connectivity and the lack of
makes it doubly necessary for non-governmental access to affordable devices continue to plague
institutions to ensure that any data in their D4D programmes in parts of the country, particularly
possession remains secure. in rural heartlands. In these environments, alternate
solutions must be provided so that no one is left
This volume showcases several approaches to behind. Anudip Foundation used a variety of
data protection. For example, Digital Green’s KDE formats to collect data, and where connectivity
initiative follows the principle of informed consent, was a challenge, they switched to manual forms.
and it is only when farmers’ express consent is Sporadic connectivity and the challenge of reading
received that their data is uploaded to the KDE app off screens led the Mindspark team to create
and further security measures applied. In Cachar, mobile audio clips and SMS-based questions which
the local District Disaster Management Authority students could respond to by typing out answers.

12
 And the ASMAN project gave healthcare Beginning to catalogue case studies of D4D during
providers SIM cards to ensure uninterrupted India’s term as G20 president assumes particular
access, and created an offline version of its app so significance because for the first time in the group’s
that data entry could continue unhindered. history, its presidency will be held by four developing
nations in succession—Indonesia, India, Brazil,
8. Work collaboratively and adopt a multi- and South Africa. With India already being hailed
stakeholder approach by many as the voice of the Global South, both
within the G20 and beyond, there could not be
D4D projects are almost always intensely a more consequential time to start mainstreaming
collaborative, and partnerships must be built at the D4D approach into the international
various levels and across diverse stakeholder development agenda. Micro Matters could serve
groups. A direct correspondence exists between as a step in this direction, and guide policymakers
multistakeholderism, scale, and success. Reliance and practitioners alike.
Foundation’s disaster management initiative in
Assam, for instance, involves sustained For Reliance Foundation and Observer Research
cooperation with the Cachar DDMA, other CSOs, Foundation, this book is the third in a series
Community Resource and Facilitation Centres, about transformations in development and
and the Inter-Agency Group (a heterogeneous governance. The first two focused on narratives
group of stakeholders that undertakes of women-led change at the community level.
humanitarian activities in the region) For PATH’s The present volume marks a shift by exploring
TB programme too, the journey from city to the interventions of organisations rather than
state to nation has been marked by an array individuals. The publishers will continue to promote
of partnerships with governments, the private research into progressive models of engagement
sector and civil society. and action, and build platforms that celebrate
personal and institutional leadership. These stories
The practices that this volume describes are of innovation and triumph must be brought to
all replicable in other settings. The methods used the widest possible audience, and the publishers
by a specific organisation and the lessons learned remain committed to doing so.
from its work could be adapted elsewhere, especially
in countries of the Global South.

- Jayashree B, Anirban Sarma, Vanita Sharma, and Shoba Suri

13
ENDNOTES

1
“‘Data for development’ will be integral part of overall theme of India’s G20 Presidency: PM Modi”, Outlook,
November 16, 2022, https://www.outlookindia.com/national/-data-for-development-will-be-integral-part-of-
overall-theme-of-india-s-g-20-presidency-pm-modi-news-237773
2
The White House, “G20 Bali Leaders’ Declaration”, November 15-16, 2022, https://www.whitehouse.gov/briefing-
room/statements-releases/2022/11/16/g20-bali-leaders-declaration/
3
Samir Saran and Anirban Sarma, “India will prioritise data for development at G20”, The Hindustan
Times, December 14, 2022, https://www.hindustantimes.com/opinion/india-at-g20-will-herald-data-for-
development-101671022759130.html
4
Common Service Centres Scheme, “Digital India”, Ministry of Electronics and IT, Government of India, https://csc.
gov.in/digitalIndia
5
Government of India, “NITI Aayog launches the National Data and Analytics Platform”, May 13, 2022, https://www.
pib.gov.in/PressReleasePage.aspx?PRID=1825145#
6
National Data and Analytics Platform, https://ndap.niti.gov.in/
7
Government of India, ‘National Data Governance Framework Policy”, July 27, 2022, https://pib.gov.in/
PressReleasePage.aspx?PRID=1845318#
8
“Will collect ‘uniform and harmonised’ anonymised datasets: Rajeev Chandrasekhar”, The Indian Express,
December 14, 2022, https://indianexpress.com/article/business/will-collect-uniform-harmonised-anonymised-
datasets-chandrasekhar-8323149/
9
Charanya Lakshmikumaran, “Digital Personal Data Protection Bill: What rights does it give individuals?’, The
Economic Times, January 13, 2023, https://economictimes.indiatimes.com/wealth/legal/will/digital-personal-
data-protection-bill-what-rights-does-it-give-individuals/articleshow/96535688.cms
10
BRICS Forum on Big Data for Sustainable Development, https://bricsbd4sd.en.cbas.ac.cn/
11
World Economic Forum, “Drones to create $100bn GDP boost and millions of jobs in India, says new World
Economic Forum report”, October 18, 2022, https://www.weforum.org/press/2022/10/drones-to-create-100bn-
gdp-boost-and-millions-of-jobs-in-india-says-new-world-economic-forum-report/
12
United Nations, “SDG India Index and Dashboard 2020-21”, https://india.un.org/en/173167-sdg-india-index-and-
dashboard-2020-21
13
Prime Minister’s Office, “PM addresses United Nations World Geospatial International Congress”, October 11, 2022,
https://www.pmindia.gov.in/en/news_updates/pm-addresses-united-nations-world-geospatial-international-
congress/
14
Priyali Prakash, “All about the National Geospatial Policy 2022”, The Hindu, January 10, 2023, https://www.
thehindu.com/news/national/explained-all-about-the-national-geospatial-policy-2022/article66346317.ece
15
Klaus Schwab, The Fourth Industrial Revolution (Geneva: World Economic Forum, 2016), p.8.
16
The Fourth Industrial Revolution, p.15.
17
NITI Aayog, National Strategy for Artificial Intelligence, June 2018, https://niti.gov.in/sites/default/files/2019-01/
NationalStrategy-for-AI-Discussion-Paper.pdf
18
Danny D’Cruze, “‘Make AI work for India’: Union Budget 2023 introduces big plans for artificial intelligence”,
Business Today, February 01, 2023, https://www.businesstoday.in/union-budget/story/make-ai-work-for-india-
union-budget-2023-introduces-big-plans-for-artificial-intelligence-368428-2023-02-01
19
Supreme Court Observer, “Fundamental Right to Privacy”, https://www.scobserver.in/cases/puttaswamy-v-
union-of-india-fundamental-right-to-privacy-case-background/#

14
16
DATA-INFORMED DECISIONS:
THE ‘ASMAN’ INTERVENTION
FOR BETTER MATERNAL
AND NEWBORN CARE

Ketaki Hate

M aternal and child health are critical aspects

of healthcare that have a significant impact on the
Foundation, Tata Trusts, and United Stated Agency
for International Development.
survival and wellbeing of women, newborns, and
children. Providing quality care during pregnancy, The ideation process for ASMAN coincided with
childbirth, and the postpartum period is essential in the introduction of the Sustainable Development
reducing maternal and neonatal mortality rates, as Goals (SDGs) in 2015, to focus on building upon
well as improving the overall health of mothers and the Indian government’s existing efforts to
their children. Despite significant advancements reduce maternal and infant mortality. The third
in the field of maternal and child health, there are SDG encourages global efforts towards ensuring
still many challenges that need to be addressed, healthy lives and promoting well-being for all. The
including access to healthcare, inadequate first target of SDG-3 concerns reducing the global
staffing, inadequate resources, and quality of maternal mortality ratio to less than 70 per 100,000
care. In 2021, India’s fertility rate was two births births and an increase in the proportion of births
per woman.1 According to the National Family attended by skilled health personnel. Additionally,
Health Survey-5, 88.6 percent of all births in the one indicator of the second target of SDG-3 is to
country are institutional births (births occurring in reduce neonatal mortality to at least as low as 12 per
health facilities), with 89 percent of all births being 1,000 live births.4
attended to by a healthcare personnel. In 2020, a

India witnessed 23,800 maternal deaths, with a In 2015, Bihar, Uttar Pradesh, Rajasthan, and
maternal mortality rate of 99 per 100,000 live Madhya Pradesh were the four most acutely
births.2 This combined with the fact that 62 percent affected states in terms of maternal mortality
of neonatal deaths occur in the first three days after rate, neonatal mortality rate, and infant mortality
birth3 means that the initial 48 to 60 hours after rate.5 ASMAN’s goal was to provide training and
delivery are critical. It is in this scenario that the mentorship to healthcare providers in public facilities
Alliance for Saving Mothers and Newborns (ASMAN) and implement technology-based solutions to
was formed in 2015, with the aim of reducing infant, improve the quality of care across eight districts in
neonatal, and maternal mortality rates. ASMAN Rajasthan and Madhya Pradesh. ASMAN’s mission
partner organisations included the Bill & Melinda was to utilise the combined knowledge and skills
Gates Foundation, MSD for Mothers, Reliance of its partners to develop tailored solutions for each

a
In addition to doctors and nurses, the NFHS considers auxiliary nurse midwives, lady health visitors, and midwives as
healthcare personnel.

17
state’s needs. In alignment with the government’s utilised technology to streamline the process
goals to decrease mortality rates, this unique of patient care and reduce the time spent on
partnership was designed to assist in implementing recording and reporting case details. It was
effective interventions in healthcare facilities. In implemented from June 2017 to May 2020,
this context, it is important to note that encouraging across 81 high delivery load public health facilities
and promoting effective public-private and civil across four districts in the states of Rajasthan
society partnerships, and building on the experience (Ajmer, Bhilwara, Kota, and Jhalawar) and
and resourcing strategies of partnership is one of Madhya Pradesh (Jabalpur, Khargone, Ratlam,
the targets of SDG-17.b and Vidisha). These facilities essentially included
sites that catered to pregnant women from low-
The ASMAN intervention is a form of Computerised income households, such as district hospitals,
Clinical Decision Support (CDSS), a digital sub-district hospitals, community health centres,
information system designed to improve clinical and primary health centres.
decision-making by healthcare providers. 6
It

LEVERAGING DATA FOR DEVELOPMENT

In 2015, when the Alliance was formed, the maternal and reports, e-learning, ASMAN game, safe delivery
mortality rate per 100,000 live births was 199 app, and remote support centres.8 The app allowed
in Rajasthan and 173 Madhya Pradesh, while the healthcare providers to record and report case
neonatal mortality rates were 38 per 100,000 births details in real-time, and to access digital registers
in Rajasthan and 32 per 100,000 births in Madhya and reports from admission to discharge during
Pradesh. 7
The intervention was launched in 2017 labour. Data confidentiality was observed, and data
and was built around the tablet-based ASMAN app is owned by the government.
and its components—case management, dashboard

b
SDG-17: “Strengthen the means of implementation and revitalize the Global partnership for Sustainable Development”.

18
ASMAN—pegged on three features, namely facilitating The app dashboard presented a visual display of key
the capacity-building of healthcare providers, utilising performance indicators in real-time, helping district
technology-based solutions, and real-time monitoring and state officials in data-driven decision-making. The
to enable prompt decision-making—aimed to use dashboard allows programme managers, government
data to save lives by improving maternal and newborn officials, and other important stakeholders to track
healthcare during the critical first 48-60 hours after patterns, recognise problems, and act promptly. It
childbirth. plays a crucial part in making data more accessible and
understandable, which leads to improved decision-
The capacity-strengthening component of the making at the policy level and increased transparency
programme sought to empower healthcare and accountability among all stakeholders involved.
providers with the knowledge and skills related to
best practices in maternal and newborn healthcare. ASMAN had three categories of beneficiaries—
Providers were trained on the Indian government’s pregnant women and their families, the hospital
Dakshata module, and post-training, mentoring, and staff being trained to use the application, and the
support visits were conducted to assist them. healthcare facilities that were being strengthened
using technology-based solutions. All three groups
As a CDSS tool, ASMAN was designed to enable of beneficiaries faced extreme stress during the
providers to make informed and evidence-based COVID-19 pandemic due to the many demands placed
decisions. The app contained digitised maternity on the healthcare infrastructure and healthcare
case sheets, e-partographs, and digital registers and providers, often at the cost of pregnant women.
reports. It enabled healthcare providers to make direct However, the ASMAN app unexpectedly created new
voice calls to a remote support centre that provided opportunities for engagement; although the app was
round-the-clock specialised guidance and decision- not designed for dissemination of information related
making support (provided by specialist doctors to pandemics, COVID-19-related protocols were built
from medical colleges) in managing complicated into the app at an accelerated pace to facilitate better
cases. This helped to ensure continuity of care and communication between government agencies and
initiation of correct treatment without delays during healthcare providers and facilities.
the intrapartum and immediate postpartum period.
The programme also incorporated gamification Another challenge for healthcare providers was
for learning via the ASMAN Game, a simulated technological, including their limited experience with
peripartum case scenario game, and an e-learning technology. Internet connectivity issues and freezing
platform hosting interactive tools, such as training tablets acted as a hindrance along with data loss
modules and the Safe Delivery app. These tools caused by bugs in the ASMAN game.10 Infrastructural
encouraged self-learning among healthcare providers challenges related to inadequate power supply
and allowed them to learn at their own pace without and low internet connectivity often hampered the
structured lessons. Describing the utility of the app, adoption of the ASMAN app. To combat this, ASMAN
Ashwini Sarathe, staff nurse at a district hospital gave the healthcare providers SIM cards to ensure
in Vidisha, Madhya Pradesh, said, “Constant use of uninterrupted internet access. Moreover, an offline
ASMAN app and its e-learning features has increased version of the app was made available to facilitate
our confidence and allows us more time for patient continued data entry and backup. Healthcare
care. The training modules include a variety of diverse providers had limited digital literacy and experience in
scenarios, helping us understand and inculcate best the use of technology. These were addressed through
practices in the management of different cases. We one-on-one sessions with trainers and grievance
are now able to give focused attention to mothers and redressals though text messages.
newborns.”9

19
IMPACTS AND OUTCOMES

The ASMAN intervention has had an impact on management of the third stage of labour, routine
healthcare systems in 81 high delivery-load public newborn care, postpartum monitoring, and
health facilities across eight districts in Rajasthan and complicated deliveries, facility readiness, availability
Madhya Pradesh. An external process of monitoring of supplies, and labour room environment. The
and evaluating the programme from 2017 to 2020 in ASMAN intervention allowed healthcare providers
Rajasthan and from 2018 to 2019 in Madhya Pradesh to make better clinical decisions due to the easy
found that there was a definite improvement in quality availability of the mothers’ medical history. The
of care. In Rajasthan, healthcare providers showed
11
training and technology implemented through the
increased knowledge regarding initial assessment intervention resulted in stronger intra-partum and
of labour, abdominal examination grips, and danger postpartum care at ASMAN facilities, resulting in
signs during the hospitalisation of mothers, along better health outcomes for mothers and newborns.
with knowledge surrounding routine newborn care. During the period in which the programme was
In Madhya Pradesh, it was found that nurses trained implemented, there was a decrease in the number
to use the ASMAN app performed better with of fresh stillbirths that occurred in the 81 healthcare
regards to knowledge and skills of maternal and facilities. From April 2019 to June 2020, the rate
newborn care when compared to nurses with no of fresh stillbirths per 1,000 live births decreased
knowledge of the app. from 7.0 to 6.4 in Madhya Pradesh, and from 7.6 to
6.7 in Rajasthan.12
Similarly, positive trends were observed regarding
practices during normal birth and labour, active

Figure 1: Rate of fresh stillbirths per 1,000 live births

April 2019 to June 2020

7.0 7.6
to 6.4 to 6.7

Madhya Pradesh Rajasthan

Source: ASMAN13

20
The success of the programme has encouraged ASMAN’s data-driven decision-making allowed
scaling beyond Rajasthan and Madhya Pradesh. for recognition of birth complications, and timely
Scalability and sustainability were designed into interventions resulted in improved care provided to
the very fabric of the ASMAN programme to ensure mothers and newborns. Dr Deepak Paldiya, Block
its longevity. A plan for transitioning the initiative Medical Officer from a community health centre
was created to guarantee its continued success in Madhya Pradesh, said, “Our facility has really
and sustainability. This plan includes building state benefited from the e-partograph - eliminating the
capacities by creating a pool of master trainers for previous trend of manual plotting. The monitoring of
clinical skills training and the ASMAN app, facilitating key vitals has improved, leading to better maternal
knowledge transfer to states, training state IT cells and neonatal outcomes. This has impacted our
on ASMAN app management, and transitioning the referral rate, which has reduced from 60-70 per
ASMAN app and database to states. Moreover, the month to 25-30 per month. The facility has
ASMAN app is also designed to be easily integrated gained newfound respect and appreciation that is
with the Indian government’s maternal health heartening.”14 A timely recognition of complications
online portals such as MCTS (Mother Child is important because a third of maternal
Tracking System), PCTS (Pregnancy, Child Tracking deaths, almost half of all stillbirths, and a quarter
& Health Services Management System), RCH of neonatal deaths occur due to complications
(Reproductive and Child Health) and SNCU during labour, childbirth, or the immediate
(Special Newborn Care Unit, the national newborn postpartum period. 15
By efficiently storing and
online portal). The goal of the process is to aid tracking data related to to the health of pregnant
state governments in assuming responsibility for women, ASMAN reduced pregnancy-related
and expanding the initiative. The project’s complications and saved time, thereby saving the
capacity-building element ensures longevity, lives of many mothers and their newborns.
replication, and scale.

21
KEY LESSONS
• Health outcomes can be improved by strengthening first-point entry facilities, such as Community Health
Centres by training healthcare providers to collect and utilise health-related data and using technology-
based solutions.
• Technology-based solutions such as the ASMAN app streamline access to healthcare, save time, and improve
clinical decision-making, thereby saving the lives of many mothers and newborns. If scaled and applied to
other health outcomes, this data-driven approach has the potential to save numerous lives.
• Before implementing technology-based solutions, it is important to train future users and provide resources
to overcome limited digital literacy and infrastructural barriers.
• Investment in public-private partnerships plays an important role in utilising institutional strength to leverage
data-driven approaches and technology-based solutions to achieve improvement in health-based outcomes
like reducing infant, neonatal, and maternal mortality rates.
ENDNOTES

1
Ministry of Health and Family Welfare, Government of India, National Family Health Survey-5 (2019-21),
http://rchiips.org/nfhs/NFHS-5_FCTS/India.pdf .
2
C Meh et al., “Trends in Maternal Mortality in India over Two Decades in Nationally Representative Surveys,” BJOG:
An International Journal of Obstetrics & Gynaecology 129, no. 4 (2022), https://doi.org/10.1111/1471-0528.16888.
3
ASMAN, Leveraging Technology for Improved Quality of Maternal and Newborn Health, 2022.
4
United Nations, “The 17 Goals”. https://sdgs.un.org/goals
5
Tata Trusts, https://www.tatatrusts.org/article/inside/five-majors-announce-project-asman-to-reduce-maternal-
and-newborn-mortality-in-india, 2015.
6
Gulnoza Usmanova et al., “The Role of Digital Clinical Decision Support Tool in Improving Quality of Intrapartum
and Postpartum Care: Experiences from Two States of India,” BMC Pregnancy and Childbirth 21, no. 1 (7 April
2021), https://doi.org/10.1186/s12884-021-03710-y.
7
Leveraging Technology for Improved Quality of Maternal and Newborn Health
8
Gulnoza Usmanova et al., “Acceptability and Barriers to Use of the ASMAN Provider-Facing Electronic Platform
for Peripartum Care in Public Facilities in Madhya Pradesh and Rajasthan, India: A Qualitative Study Using the
Technology Acceptance Model-3,” International Journal of Environmental Research and Public Health 17, no. 22
(2020), https://doi.org/10.3390/ijerph17228333.
9
Quote originally from Leveraging Technology for Improved Quality of Maternal and Newborn Health.
10
Usmanova et al., “Acceptability and Barriers to Use of the ASMAN Provider-Facing Electronic Platform for
Peripartum Care”
11
Leveraging Technology for Improved Quality of Maternal and Newborn Health.
12
Leveraging Technology for Improved Quality of Maternal and Newborn Health.
13
Leveraging Technology for Improved Quality of Maternal and Newborn Health.
14
Quote originally from Leveraging Technology for Improved Quality of Maternal and Newborn Health.
15
Usmanova et al., “The Role of Digital Clinical Decision Support Tool”

23
24
‘COTTONACE’:
WEAVING AI INTO AGRICULTURE

Shimona Mohan

I ndian cotton has aptly earned itself the moniker

‘white gold.’ Cotton has significant commercial and
global average of 768 kg/ha,4 because approximately
30 percent of crops are lost every year to hard-to-
cultural importance for the country—India has the detect pests like pink bollworms (PBW) and American
highest cotton acreage in the world and accounts bollworms (ABW).5 For smallholder farmers, this
for around 25 percent of the total global cotton results in a significant loss of income and leads to the
production.1 Indian cotton has also been recognised indiscriminate and often ineffective use of pesticides.
as the healthiest fabric, and India remains the only
country to cultivate all four species of cotton on a Wadhwani AI, an independent non-profit institute
commercial scale.2 These statistics are all feathers in developing AI-based solutions for underserved
the cap of the estimated six million farmers in India communities in developing countries,6 took
whose livelihoods revolve around cotton cultivation. 3
cognisance of this issue in 2018 when the
Maharashtra government brought it to their
However, growing cotton comes with its own set of attention. The institute spent the next few years
challenges for a majority of these farmers who are developing the CottonAce application, an AI-
smallholders (those who own and cultivate less than powered early warning system that can be
two hectares of land) or are landless (those who downloaded on a smartphone, to help smallholder
are simply agricultural labourers but do not actually cotton farmers protect their crops from pests
own any land), and thus lack the resources to adopt like PBW and ABW. When the app was piloted in
sophisticated pest management measures. Despite 2021, it was used by nearly 6,000 farmers and
the high acreage, the average cotton yield in India extension officers across 10 cotton-growing statesa
is 451 kg/ha, which is significantly lower than the and 60 districts in India.

LEVERAGING DATA FOR DEVELOPMENT

While CottonAce is free to download in nine and the resources to buy pest traps. The lead
vernacular languages, it works on a registration farmers are recognised and registered on the app
model for ‘lead farmers’, who may be landowners by Wadhwani AI’s local partner organisations and
with access to a smartphone, an internet connection, work closely with farmer welfare programmes run

a
The 10 states are: Andhra Pradesh, Gujarat, Karnataka, Madhya Pradesh, Maharashtra, Odisha, Punjab, Rajasthan, Tamil
Nadu, and Telangana

25
by the latter to manage pest infestations in the areas from the lead farmers. The lead farmers upload
they service. Another category of farmers, ‘cascade images of pests from the farm onto CottonAce, and
farmers’, may not have the app or the required traps, the app issues contextual alerts and/or advisories on
but receive alerts and advisories about the pests recommended actions.

Figure 1: Village-level dissemination of advisories used by CottonAce

ACTION
TAKEN
Based on
advisory

ADVISORY
GENERATED
Localised, real-time,
accurate, verifiable
LEAD FARMER EXTENSION CASCADE FARMERS
Monitors infestation OFFICER Benefit from advisories even
using the app Verifies alerts without smartphone access
and disseminates
advisories

COTTONACE Al
Localised, real-time,
accurate, verifiable

CottonAce uses both primary and secondary these errors as well. Since CottonAce uses an AI-
data to generate insights for its users. During the based solution that is built into the app, it does
initial phases, Wadhwani AI’s partner organisations not need a constant internet connection. It only
provided relevant information and meta data that requires an internet connection when the app
was needed to lay the groundwork for the app. needs to be downloaded and installed.
After it became functional, primary data has been
collected from the farmers’ use of CottonAce, and • Secondary data: This is localised data provided
the partner organisations have been providing by Wadhwani AI’s partner organisations, who run
secondary data for backend operations. The data the farmer welfare programmes. They maintain
encompassed under each of these categories is as information regarding the farms, extension
follows: officers, and farmers under their purview. This data
is used solely to facilitate the private web-based
• Primary data: This comprises images of pests administration portal used by the programme
caught in the traps that are taken and uploaded administrators to monitor the usage of the app
by cotton farmers in the course of using the and the level of infestation over the duration of
CottonAce app. The AI used by CottonAce the season.
has been trained on about 40,000 images and
functions on a detection and rejection machine In all phases of data collection and use, Wadhwani
learning (ML) model for pest types. It can also AI is conscious of data privacy and the responsible
recognise when the image is not properly taken or use of AI. “We ensure that no personal data or
the insect is not clearly visible, and recommends metadata is used for our machine learning model,
that farmers take another picture in such a and that responsible AI is used with the involvement
situation. If the farmers feel that the insects have of both internal and external experts so that no
been incorrectly identified, they have the option community is put at risk,” said a member of the
to report this on the app and the AI learns from Wadhwani AI team engaged with the technical

26
aspects of CottonAce. The institute also follows agricultural advisory taking into account when
this by signing non-disclosure agreements with the farmer last sprayed pesticide, which kind was
its partner organisations and ensuring that they used, for what purpose, and how much. This is
follow compliance-related measures during the particularly useful for farmers who would otherwise
collection and processing of farmers’ personal have to wait about 10-15 days for their regional
data. While CottonAce receives some personal agricultural departments to issue an advisory, or
information (such as names and phone numbers) who would need to call the Kisaan (farmer) call
when the farmers install the app, no other information centre for advice. Faster advisories are required
is required, and the app’s data servers are located in in agricultural timelines, and with CottonAce,
India itself to ensure data sovereignty. farmers are constantly informed and do not need
to depend on anyone else to take quick and critical
After the pictures of the pests are uploaded actions to preserve their cotton crops.
on CottonAce, the app instantly generates an

Figure 2: Working of the CottonAce app

No Action
Detection

Input Action Wait and

Counting Count
Threshold Watch

Shared Base Spray

Yes
Network Image Validation Pesticide

Trap? No Retake Image

While CottonAce is a novel solution for any and are willing to try it since working with the app
cotton farmer seeking to improve their yield, results in time conservation and money savings.
Wadhwani AI initially had to persuade farmers to The team also feels that younger farmers are more
change their traditional way of working. “It was receptive to new tech interventions. About 1,500
hard to convince them that their efforts and our farmers learnt about the CottonAce app through
efforts are for the mutual goal of the welfare of the the grapevine and downloaded it, even though they
community. Not everyone understands AI – without could not use the app since they needed to be
data, we have no solutions, and without solutions, registered through the backend by Wadhwani AI’s
we have no data to build further on,” said a member partner organisations. The institute is currently
of the CottonAce execution team. working on new rollout plans to fix this so that any
interested cotton farmer across India can use the
However, CottonAce has gradually overcome service, even if they are based where there is no
this. Local communities are no longer as resistant partner organisation.

27
IMPACTS AND OUTCOMES

Wadhwani AI regularly collects feedback based on The last principle is an organisational mandate
an exhaustive questionnaire of over 100 questions for the institute. Wadhwani AI also believes in the
on performance, which is analysed by a dedicated positive impact of open-source datasets and
team. The partner organisations then increase or technologies on development endeavours, and have
reduce monitoring and initiate course correction put this into practice by releasing an open-source
based on the information provided to them at scaffolding of their ML model to which anyone
the end of every monsoon season (June-October). can contribute. The institute’s efforts and results
Over the past two years, CottonAce has worked with CottonAce have made state agriculture
with 15 partners, 30,000 cotton farmers (of which departments aware of how AI can be used for
10,000 are lead farmers, and 20,000 cascade their benefit. Wadhwani AI signed a memorandum
farmers) spread over 140 talukas over 30 districts in of understanding with the Indian government in
all 11 cotton-growing states. However, the COVID-19
b
May 2022, and have also reached out to and
pandemic has slowed things down. “Our numbers collaborated with research organisations working
could have been higher, the adoption could have on similar projects. The institute’s experience with
been faster, and the feedback we received could have CottonAce has also motivated them to delve into
been stronger,” said a member of Wadhwani AI team similar services for other crops, the foundations for
involved with CottonAce. which are underway.

Despite the challenges, Wadhwani AI has retained the Amidst new developments, a few factors that
three principles on which CottonAce is based: have helped them bring AI to the fore in terms
of agriculture and build CottonAce into a useful
• Improving the efficiency of the cotton cultivation development for cotton farmers in India are: building
system; trust, generating comprehensive solutions, and
• Ensuring the farmers get context-specific sharing lessons with relevant stakeholders.
advisories that they can use to take informed
decisions; and
• Helping ecosystems understand how AI works for
them.

b
The 11 states are: Haryana, Punjab, Rajasthan, Gujarat, Maharashtra, Karnataka, Tamil Nadu, Andhra Pradesh, Telangana,
Madhya Pradesh, and Odisha

28
KEY LESSONS
• It is essential to help multifarious communities and professions understand how data and AI can be used in
consonance with their existing ways of working to help them improve their efficiency.
• Updating and upgrading any data-based solution according to sectoral specificities should be a regular
practice.
• Ethical and responsible use of data and AI is crucial in any scalable model that uses data for development.
ENDNOTES

1
Ministry of Textiles, Government of India, Annexure VII – Cotton Sector, 2022, https://www.texmin.nic.in/sites/
default/files/Cotton%20Sector.pdf
2
Ashima Sharda Mahindra, “India at 75: The Humble Indian Cotton Is the Healthiest Fabric Ever,” Times Now, August
12, 2022, https://www.timesnownews.com/health/india-at-75-the-humble-indian-cotton-is-the-healthiest-fabric-
ever-article-93510613
3
Ministry of Textiles, Government of India, Annexure VII – Cotton Sector, 2022, https://www.texmin.nic.in/sites/
default/files/Cotton%20Sector.pdf
4
Ministry of Agriculture & Farmers’ Welfare, Government of India, Annual Report 2020-2021, 2021, https://agricoop.
nic.in/sites/default/files/Web%20copy%20of%20AR%20%28Eng%29_7.pdf
5
Babasaheb B. Fand et al, “Degree day-based model predicts pink bollworm phenology across geographical
locations of subtropics and semi-arid tropics of India,” Scientific Reports 11, 436 (2021), https://doi.org/10.1038/
s41598-020-80184-6.
6
Wadhwani AI, https://www.wadhwaniai.org/

30
32
E-SOLUTIONS AS A WAY TO
EMPOWER COMMUNITIES
WITH DATA
Sitara Srinivas

O f the 43 percent of Indians with access to the

internet, a smaller percentage is female (only 15
1
greater female decision-making power, shape role
models who can empower others, and help create a
percent of the total female population have access transformed community with a reduced digital gender
to the internet). The National Family Health Survey-5 divide. Data was the bedrock of this project; Anudip
(2019-21) highlighted the significant gender gap in collected data first to survey the field, and then used
internet usage, suggesting that only 57.1 percent of this data for targeted development interventions.
the male population and 33.3 percent of the female
population had ever used the internet.2 Seen through Using their proprietary Computer Management
the urban-rural lens, this gap becomes even more Information System (CMIS) and Learning Management
prominent, with 51.1 percent of urban females and System (LMS), Anudip worked with over 75,000
only 24.6 percent of rural females qualifying as having women across five states (West Bengal, Jharkhand,
used the internet. Demographic identities like caste,
3 Odisha, Assam, and Rajasthan). These states were
class, and religion could further skew this gap. picked because of their high (and, thus, troubling)
gender divide.a According to Narendrayash (Naren)
Today, with the world accessible through a from the Anudip programme management team, “The
phone screen, the internet empowers many. But, whole project was localised in terms of mobilisation
unfortunately, given issues around access, this and how training was conducted, but on reach, we
empowerment remains a challenge to many. wanted to reach as deep and far as possible. That
was our concept. How far can we reach, and how
Anudip Foundation, a Kolkata-based organisation, many lives can we impact? The impact we have – if
“creates digital livelihoods for underserved it is on scale, the whole community benefits from it.
communities through technology and skills”. Through And initial results have shown us that some of these
a project launched in 2021, Anudip aimed to help changes have happened.”
bridge the digital gender divide in India to encourage

a
Each state also had additional distinctive reasons. For instance, Assam was picked since working there would help
create an impact in India’s Northeast. Odisha and Jharkhand both have a high level of community-based marginalisation.
Rajasthan has some of the lowest gender parameters overall. West Bengal is both Anudip’s home and a state with some
of the lowest socioeconomic parameters in certain rural pockets.

33
LEVERAGING DATA FOR DEVELOPMENT

Anudip is heavily reliant on three main software Data privacy, confidentiality, and security were crucial
packages for its work. The first is their CMIS system, to Anudip throughout the process. To maintain the
which acts as a repository of all its beneficiaries and sanctity of the data, the organisation has instituted
as a tracking, analysing, and monitoring system. Bittu parameters for data privacy and protection that all
Adikari, cluster head at Anudip, explained, “Since we employees need to maintain. The CMIS is locked, with
are working with a big set of numbers, the CMIS was each employee being given credentials for only their
an essential piece of the data collection process.” projects, ensuring that only those authorised receive
The second piece of software is the LMS, a training access. For those working for or with Anudip on the
and monitoring software that direct beneficiaries field, the LMS included training modules on data
can access. Through the LMS, Anudip also conducts hygiene, with everyone being made to take the data
assessments of the direct beneficiaries and tracks oath.c To ensure that people were made aware of cyber
individual scores, the percentage of modules fraud in an accessible way, Anudip produced posters
completed, and other insights. The final piece of visualising safe internet practices. Additionally, data is
software is Google Forms (Anudip has a licensing not stored, only recorded through Google Forms, with
agreement with the company), used to collect only the project team able to access this data.
baseline and inline data processes.
Anudip uses the data collected to enrol these women
About 24 community mobilisers work through Anudip in governmental insurance schemes, as sellers on
with panchayat and self-help groups to connect with
b
e-commerce platforms, and as beneficiaries of other
communities. Within these communities, Anudip government schemes. Once registered, Anudip stores
facilitators train direct beneficiaries on how to use screenshots and confirmation proofs to continue to
Google Forms. Data (names, ages, addresses, other monitor that the women receive funds.
contact details, and accessibility and availability of
phones) is then collected and stored temporarily Anudip also analyses the data to study demographic
in a Google response sheet, with a validation test distribution in their work regions. In particular, two
being conducted to check for any errors. Once the critical transitions among the women in the area
information is checked and processed, it is added to from pre- to post-enrolment have been tracked, the
the CMIS database. first being phone ownership (as many women do
not have their own or are allowed access to their
Anudip works directly with women as data collectors, phones) and the second being changes in occupation
not taking on the service of any other third party and income. Changes in behaviour due to increased
or NGO. Without the help of software, this reliance incomes, including the propensity of social media
would be tricky. In empowering others, data is helping usage and search activity on web browsers,d were
women empower themselves. also observed. Finally, Anudip tracked women’s role

b
The panchayat, composed of the words panch (five) and ayat (assembly), is the group of five that govern at the village
level in India.
c
Similar to the Indian government Cyber Awareness Security pledge, Anudip’s cyber oath lists and commits to the basic
safeguards and measures one can take to avoid cyber fraud online.
d
For instance, what were people looking for online.

34
in their households’ decision-making structure to see and introduce platforms such as Haqdarshak.f Adikari
if there had been any changes post their intervention. explained, “Education has been done on the ground,
and now people can ask for their rights from the
In Haripur, a village in Odisha about 40 kilometres government”. During COVID-19, when people faced
from the nearest township, Anudip witnessed how problems accessing funds, Anudip trained them on
women were not allowed access to phones or social digital banking. Anudip also established connections
media. After convincing the village panchayat, Anudip with popular e-commerce sites and ensured last-mile
held a seminar for the women in the village (and the delivery connectivity.
men in their families), where they shared details about
the potential of the tech and data. Slowly, women In a survey, Anudip asked the women they worked with
here began using social media. Jasminara Begum, a if their position in the household had changed. Over
single mother who sold saris, would earlier go door to 80 percent responded yes. Community members now
door from village to village to sell her saris. Today, she reach out to these women for mobile phone recharges
sells saris sitting at home through Facebook Live and and other transactions. As such, the involvement of
receives payments through digital platforms. Because women in the community has grown tremendously.
of Haripur’s off-grid location, villagers would spend a
day going to the closest bank for even the most minor Innovativeness, challenges, and solutions
transactions. Through Anudip, Mandakini opened a
Customer Service Point (CSP)e in collaboration with The project’s innovativeness emerged from three
Canara Bank. This CSP now caters not just to Haripur key sources. The first was the diversity of the states
but also to several villages in the vicinity. Similarly, that Anudip worked with. This diversity did involve
Priyanka Debnath has been able to fund her treatment challenges, given local variances and nuances that
for thalassemia by engaging in reselling on online could often be a barrier. But with the programme
platforms. aiming to be as pervasive as possible, they worked with
local government agencies to overcome these. The
Narendrayash said, “With data, the awareness and data collected was localised by bringing in mobilisers
transformation you can bring to the whole community from those very communities. This attention to detail
through just one participant is incredible. A success helped improve data quality and accuracy. Anudip
story snowballs through the whole community also used a variety of formats to collect data. In
bringing in many more.” While there are many big places where the internet was scant, they relied on
examples, Naren also suggests that the smaller, less manual forms. Using an amalgamation of methods
heard stories tell the story of how data has impacted also helped in places where the full use of tech would
both Anudip and their beneficiaries’ world. not have worked.

Data has also brought governance closer to people. The second source was the key audience—women.
Earlier, there was a general lack of awareness and Anudip deliberately chose to work with women from
intimidation regarding knowledge of government rural pockets who had not had or usually had limited
schemes and programmes. Through data, Anudip access to data.
could identify pockets where intervention was required

e
Customer Service Points are mini banks that can be opened and run by people who live a distance from the closest
bank. It can be used for to create bank accounts as well as to deposit and withdraw money.
f
The Haqdarshak platform can be used to check the availability and eligibility for government schemes.

35
Finally, Anudip aimed to engage and change mindsets. miscommunication and misinformation; thus, they
This was at two levels—first, that women should be kept would use conventional methods instead of tech.
from tech; second, that digital platforms were unsafe.
Because of stigma and the challenges of patriarchal Often using big data erases individual issues.
norms, it was often hard to collect the required data However, because of the localised nature of Anudip’s
from women. To encourage changes in mindset and programme, they were able to engage with and
to address the apprehensions, Anudip engaged with alleviate individual issues to ensure that everyone was
local government agencies and conducted seminars able to participate in the programme entirely.
and workshops which involved both women and men
from the community. Another challenge was that For Anudip, COVID-19 was an opportunity because of
some people were adversarial towards technology. how common digital platforms became and because
They were intimidated by technology through people saw how tech benefited others in many cases.

IMPACTS AND OUTCOMES

Through this project, Anudip had a transformational the community has benefited overall. Earlier, they
impact on the lives of the people they worked with. would have to leave the village and travel distances
The change occurred not just at the individual but also to complete simple tasks. Today, they most likely just
the community level. While it was initially difficult for need to ask their neighbour for help.
Anudip to get women to work on this project, women
are now keen to work with the organisation because of For Adikari, “Data is an important pillar for development.
the programme’s impact. Over 3,000 women working Through data, you can analyse the ground realities,
as online resellers have a Facebook group where they the gaps one has to overcome, and the areas one
regularly engage and share their experiences. There needs to work in. with data, the complete scenario is
are also WhatsApp groups where women stay in made available.” For Narendrayash, “Data is the fuel
touch with each other and the foundation. As a result, of the 21st century. It is imperative to look at data for
people have been inspired to be digitally literate and any intervention. Through data, we understood the
fully embrace tech. demographics we were working with and were also
able to measure the shifts in the programme. Data
Impact surveys conducted by Anudip show the level was integral for us to understand if we were working
at which these women have been made financially in the right direction. Data has been important for
independent. Because these women have taken up any intervention in the development sector, and it will
various occupations—managing a CSP, working as a grow to be more significant.”
government scheme enroller, or even as a merchant—

36
KEY LESSONS
• To empower women, it is crucial to involve men in the household as well, and to focus on creating
mindset changes.
• Empowering women often indirectly or even directly empowers the whole community as the benefits and
advantages are often quickly spread far.
• While using data for development is useful to understand the macro picture, engaging with the micro level
is also crucial for a successful intervention.
ENDNOTES

1
ITU, World Telecommunication, “ICT Indicators Dashboard – India” https://www.itu.int/en/ITU-D/Statistics/
Documents/DDD/ddd_IND.pdf
2
Basu Chandola, “Exploring India’s Digital Divide.” ORF, May 20, 2022, https://www.orfonline.org/expert-speak/
exploring-indias-digital-divide/
3
Basu Chandola, “Exploring India’s Digital Divide.”

38
40
‘KISAN DIARY ENTERPRISES’:
A MOBILE SOLUTION FOR INDIA’S
SMALL FARM SECTOR

Saji Kadavil

O ver the last decade or so, India’s agricultural

sector has seen a change with the emergence of
need to be addressed to make FPOs a sustainable
business model.11 These include a lack of technical
digital technologies. This comes on the back of skills and market linkages, inadequate knowledge
growth in India’s digital economy. In a December 2022 about government and other grants, weak finances,
report, the Reserve Bank of India noted that insufficient professional management, and a lack of
between 2014 and 2019, the country’s digital risk mitigation mechanisms.12
economy grew 2.4 times faster than the overall
economy.1 Agriculture and allied sectors accounted In 2019, Digital Green, a development organisation
for about 16 percent of the gross value added in that works with smallholder farmers, started the
FY 2022, 2
and employed 42.3 percent of the Kisan Diary Enterprises (KDE) mobile app to help
population in 2019-20, 3
with smallholder farmers FPOs gather and share data to facilitate buyer
making up a large share (about 78 percent). collecting and group decision-making, and
Additionally, the sector has seen a robust annual aggregate farm-level data for input and output
growth rate of 4.6 percent over the last six years.4 markets. KDE has been instrumental in ensuring
inclusive data and innovative technology usage
However, India’s small-farm sector faces numerous to address some of the challenges of the FPOs,
challenges, including market uncertainties, increasing especially those represented by women in India’s
land fragmentation, and high production costs. 5 South and East. Since 59.5 percent of the 30,718
One solution is the formation, development, and active FPO members are women, the organisation
commercialisation of Farmer Producer Organisations uses the app to ensure continued inclusion among
(FPOs). FPOs are village-level associations of farmers,
6 small farmers. KDE was first implemented in Odisha,
especially small and marginal farmers, with between and was later expanded to Jharkhand, Bihar,
15-20 members, established after the amendment of Telangana, and Andhra Pradesh, covering 30 FPOs
the Companies Act in 2003. Over the years, FPOs
7 in total. All these regions are represented by the 64
have shown that they have the power to create a crop profiles on the app. During the COVID-19
better ecosystem for farming, improve economies
8 pandemic, various app-based solutions emerged
of scale, strengthen the agricultural community and for India’s agricultural space.13 The data collected by
sustain their livelihoods. 9
About 30 percent of all KDE enabled Digital Green to partner with many
FPOs are running profitably, while 50 percent are agricultural service providers and curate better
still in the development stage.10 Still, certain issues services for the farmers.a

a
For instance, in Telangana, about 30 percent to 40 percent of farmers use the KrishiTantra app services for conducting
the soil test.

41
LEVERAGING DATA FOR DEVELOPMENT

KDE involves four phases of activity to collect and use information related to cultivation of crops, details on
data for the platform. sowing, reported pests and weeds, outputs (including
crop name, area of cultivation, allocation of products
In the first phase, the organisation uses secondary for FPOs, storage for self-consumption, and estimated
data to identify the beneficiaries. The data collected harvested data), and the required inputs. Farmers are
includes the profile of FPOs, details of their members, encouraged to upload this information from their
list of dominant crops, crop intensity, cultivation smartphones. This is done with the farmers’ informed
practices and related issues, challenges related to consent.b Data is validated during FPO meetings with
managing pests and weeds, and market studies. This producer groups of the relevant FPOs in each region.
serves as a foundation for the generation of data and
for the development of service assistance features to The final phase of the process is generating a dashboard
address the major problems, particularly about crops and analytical reports for FPOs and farmers. The KDE
and geographical areas. The platform also relies on app generates a dashboard using the aggregated data
frontline workers, the livelihood missions of the state for FPO leaders/directors to leverage for collective
that are part of the National Rural Livelihood Mission, decision-making, sharing with public institutions,
and the state’s agriculture department to provide and for input and output market transactions. The
services at the farm level. aggregated data on the quantity of harvested crops
available to FPOs and the requirement for inputs like
The next level involves identifying and training seeds and fertilizers is readily available to the FPO
frontline workers on using the KDE app and the data head. Frontline workers/community resource persons
collection and operating process. Through frequent get a summary of queries from the farmers on sources,
group discussions, the frontline workers raise plant protection, and required equipment, along with
awareness about the app and the value of the data other information on final products available for sale.
platform among FPOs and members. A separate summary of practices adopted by the
farmers under each category of the production cycle
The next step is to directly approach FPOs and their is available for further assistance.
members. Data is collected on farmers’ profiles, basic

b
The frontline workers take consent from farmers before the survey.

42
Figure 1: Generation of data via KDE and its use

Secondary data for selection

of regions, FPOs, farmers

Farmers
Selection of frontline workers
and training on KDE and data

KDE Dashboard,
database analytical
reports and FPOs
Data collection and
validation from farmers videos

FPOs and farmers Private agri

apps

Partnering for ICT

services

For an analytical and interactive dashboard, to precise information on the production of a certain
disaggregated data is analysed and visually crop in a certain region, including its quantities. With
represented. The data gathering and analysis reliable information on quality and quantity, the FPOs
processes happen continuously. Data is periodically can negotiate with buyers for a more competent price.
updated in accordance with the crop season, and The KDE dashboard helps the directors make quick
is used to generate analytical reports and in-depth decisions while negotiating with buyers.
comparative analysis of crops’ financial profit and loss.
The insights are used at various levels, from decision- FPOs are also able to bargain more effectively with
making and information transmission processes at the input suppliers because of the aggregate data on farm
farmer and institutional levels, to designing projects size and crops. This new approach decreases input
funded by the state agriculture department. waste, as farmers do not spend money on making
significant purchases without having the necessary
During the COVID-19 pandemic, farmers used background information. Importantly, the KDE app
the mobile app to update information on seed has made the process of harvesting, selling and buy
requirements and harvested produce. Importantly, this produce less time consuming. FPOs get adequate time
allowed the FPOs to partially enable trade and curbed to identify suitable buyers since the data can predict
the number of distress sales during this period. the quantity of a crop that will be produced. Frontline
workers and FPOs upload information on behalf of
Data-based decisions help the FPOs to negotiate individual (or member) farmers regarding their crop,
effectively with buyers and service providers based which also facilitates easy decision-making by the
on sheer volume. The economies of scale that are buyer. Buyers have access to exact product quantities,
built into the structure of FPOs are becoming more product variety, and pricing range information.
and more apparent. The board of directors has access

43
IMPACTS AND OUTCOMES

KDE has had many positive impacts since its roll-out.c financial subsidies and credit now. Because I am
It has been instrumental in helping FPOs tackle some talking based on data, not based on a vague idea. The
of the issues that plague them, including of inadequate inputs we provide to the KDE database give us back
professional management and low-price realisation in production and productivity forecasts.” Indeed, even
input and output markets. officials from NABARD have advised other farmers to
document their farming methods in a similar way.e
The FPOs can successfully bargain with purchasers
and service providers based on aggregate data and KDE is not a project-based initiative; it generates a
analytical reports with the aid of data. The leader of large-scale live database of farming practices, financial
Maa Murgasuni,d an FPO of tribal women in Odisha’s transactions, and profiles of farmers. According to
Mayurbhanj district, noted how KDE helped her KDE’s growth pattern, it has built a sizable database
negotiate a better price for pumpkins harvested from that serves as a platform for cooperation, public-
311 marginal farmers who cultivated 391 acres of their private partnerships, and the development of small
land with buyers: “The market is generally dominated farmers into commercial farmers.
by men. Earlier, we faced a problem in negotiating
better prices without accurate data. KDE provides us The KDE mobile app has now been expanded into a
with accurate and aggregate data on total harvested broader database platform. Digital Green uses KDE
crops with the image showing the products’ quality. for some of its major agricultural projects.14 Under the
It helps us to negotiate better with buyers based ‘Saagu Baagu’ project for the chilli crop in Telangana,
on data, consistently fetching about 20-30 percent an AI chatbot is frequently used for knowledge
higher prices.” distribution based on the KDE database. The ‘Atlas’
(Advancing Tribal Livelihoods and Self Reliance)
Farmers have reported that regular data collection project in Odisha and Jharkhand for tribal women
has raised greater awareness on the importance farmers also uses the KDE database. ‘eMircha’ is
of meticulous record-keeping. An Odisha paddy another project that uses the KDE database to deliver
farmer remarked, “I am confident to meet with digital advisories to more than 31,073 chilli farmers in
officials from the National Bank of Agriculture and Andhra Pradesh and Telangana.
Rural Development (NABARD) and banks to discuss

c
KDE’s success is in part attributable to the implementation process. Ronali Pradhan, Digital Green’s operations head for
Odisha and Jharkhand, and her team played a pivotal role in facilitating systematic capacity building along KDE’s chain
of operations, and the intensive implementation of the programme across all regions.
d
Maa Murgasuni has been institutionalised by a state-level civil society organisation, Centre for Youth and Social
Development. The FPO also receives technical assistance from the Odisha Livelihoods Mission, which is part of the
National Rural Livelihood Mission.
e
As told to this author by Ronali Pradhan (Digital Green’s operations head for Odisha and Jharkhand) through her
personal communication with farmers in September 2022.

44
KEY LESSONS
• Digital technologies present innovative opportunities to ensure inclusiveness while addressing the prime
issues facing India’s smallholder farming sector.
• There is scope for live database platforms to collaborate with other agritech solutions.
• Technology upgradation and integration (the constant improvement of an app) is vital to ensure continued
assistance to farmers.
• The easy availability of aggregated data empowers women FPO leaders and marginalised farmers to
negotiate better market prices and economies of scale.
ENDNOTES

1
Reserve Bank of India, Monthly Bulletin, December 2022, Mumbai, RBI, 2022 https://www.rbi.org.in/Scripts/BS_
ViewBulletin.aspx?yr=2022&mon=12
2
Atri Mukherjee, Priyanka Bajaj, Rishabh Kumar and Jobin Sebastian, Indian Agriculture: Achievements
and Challenges, Mumbai, Reserve Bank of India, 2022, https://rbidocs.rbi.org.in/rdocs/Bulletin/
PDFs/02AR_170120226CD582783DB44FECB7A07AC238270E5F.PDF
3
Ashok Gulati and Juneja R. ‘“Transforming Indian Agriculture,” in Indian Agriculture Towards 2030, ed Chand R et.
al., (Singapore: Springer, 2022), https://doi.org/10.1007/978-981-19-0763-0_2
4
Government of India, Ministry of Finance. Economic Survey: 2022-2023 (New Delhi, Ministry of Finance, 2023),
https://www.indiabudget.gov.in/economicsurvey/
5
“Indian Agriculture: Achievements and Challenges”
6
Shiwani Bhadwal et.al., “Farmer Producer Organization: A Potent Tool for Paradigm Shift in the Farm Sector,”
Journal of Agricultural Extension Management 23, no.1 (2022). https://www.manage.gov.in/publications/journal/
jan-jun2022.pdf
7
Ministry of Agriculture & Farmers Welfare, Government of India, https://pib.gov.in/Pressreleaseshare.
aspx?PRID=1696547
8
Mourya A. and Mehta M. “Farmer Producer Company: India’s Magic Bullet to Realise Select SDGs?,” International
Journal of Rural Management, 17, no. 1 (2021) https://doi.org/10.1177/0973005221991660
9
Chandre Gowda M J, Sreenath Dixit and Megha H L, “Women’s Participation in Karnataka’s FPOs,” Economic and
Political Weekly 53(2018): 45.
10
Annapurna Neti and Richa Govil, Farmer Producer Companies: Inclusion, Capitalisation and Incubation’ (Bengaluru:
Azim Premji University, 2020). https://cdn.azimpremjiuniversity.edu.in/apuc3/media/publications/downloads/
report/Farmer-Producer-Companies-Report-2-Azim-Premji-University-2022.pdf
11
Shawn A Cole and Nilesh Fernando A, “Mobile’izing Agricultural Advice Technology Adoption Diffusion and
Sustainability,” The Economic Journal 131 (2021): 633.
12
Pingali P et. al., “Linking Farms to Markets: Reducing Transaction Costs and Enhancing Bargaining Power. In:
Transforming Food Systems for a Rising India,” Palgrave Studies in Agricultural Economics and Food Policy
(2022) https://doi.org/10.1007/978-3-030-14409-8_8
13
Anjani Kumar et al., “How Indian agriculture should change after COVID-19,” Food Security. 12
https://doi.org/10.1007/s12571-020-01063-6
14
Digital Green, Annual Report-2020 (New Delhi: Digital Green, 2021), https://www.digitalgreen.org/wp-content/
uploads/2021/05/DG_2020_AR_high_res_spreads.pdf

46
ENABLING DISASTER
PREPAREDNESS WITH
GEOSPATIAL DATA
Mohini Ganguly

In December 2022, the Indian

notified the National Geospatial Policy, which aims
government their pre-disaster conditions. Working at the grassroots
for many years, Reliance Foundation’s Disaster
to build a national framework that supports the Management programme also understands the
liberalisation and commercialisation of the sector. importance of moving towards disaster preparedness
This has foregrounded the increasing importance in addition to disaster response to mitigate the worst
of geospatial data for various purposes, including impacts of recurring flooding.
in disaster risk reduction.
After the relief efforts, Cachar District Disaster
Assam, a state in India’s Northeast region, is Management Authority (DDMA) approached
vulnerable to a host of disasters due to its Reliance Foundation to be a knowledge partner in
topography. According to the Center for Study of developing disaster management strategies. The
Science, Technology and Policy’s Climate Atlas, approach takes into account the importance of
which analyses historical data and presents near- incorporating extremely localised challenges and
future scenarios, all of Assam’s districts will see creating similarly localised responses and plans,
increasing rainfall and nearly all will see increasing not only at the district level, but at the community
high-intensity rainfall events 1
as climate change or hyperlocal level. A holistic disaster management
intensifies, under the scenarios they have analysed. approach saves lives and livelihoods. Supporting
public institutions in disaster preparedness and
Cachar district is a landlocked region, bordered by community resilience-building for disaster risk
the states of Meghalaya in the north, Manipur in the reduction are key paths to reducing losses when
east, and Mizoram in the south. The Barak River runs disasters strike. Supporting the community and
through the district and, as such, it is vulnerable to the local government in identifying key areas of
flooding, landslides, and earthquakes. During such vulnerability and developing tools and gathering
disasters, communication becomes challenging, data to make decisions quickly during disasters
heightening the area’s vulnerability. are important aspects of holistic disaster
management. Reliance Foundation’s approach to
In May and June 2022, Assam experienced this programme is also built on these pillars. Using
extensive floods, with over 55 lakh people impacted.2 a Geographic Information System (GIS), it aims to
Reliance Foundation engaged in relief operations use geospatial data for facilitation of disaster
in the Cachar and Nagaon districts, with the planning and preparedness by decision-makers.
assistance of volunteers to support the district While the district’s database contained GIS
administration and NGOs, benefiting approximately based maps, and infrastructure related data, it
30,000 people. Disasters can have a multigenerational required support in developing methods to use
impact, and it can take years for families to recover to these resources systematically, and identifying and

49
addressing gaps. For instance, support in reducing for providing early warning advisories in
information chasms, or the lack of mechanisms floodprone areas.

LEVERAGING DATA FOR DEVELOPMENT

A GIS can have numerous uses in all phases of human environment. The approach being used in
disaster management (mitigation, preparedness, Cachar brings together both facets, using household-
response, and recovery). In respect to disaster level surveys and critical infrastructure-mapping
management, risk is now understood to be a to identify human environment risks, and likelihood of
function of both the geography of an area and the flooding or other hazards to identify geography risks.

IMAGE 1: Process of Creating a District Hazard Risk Map

Non-Spatial Data Spatial Data

Historical event Administrative Digital River, Rail, road, Household mapping/POI

disaster data
Data (Flood/Wind/
boundary (State, Elevation water population facility mapping Base map/
District, Taluka, Village) Model bodies density (school, anganwadi, topographic map/
point Cyclone hazard police station, hospital,
atlas, flood layer) satellite map
community hall etc.)

Link the flood hazard Route network /

Historical Verify low Hydrology demographic
Purpose information classification with lying area Critical Infrastructure
village boundary network Graphic
mapping representation of
NRSC, Bhuvan, earth surface
Source Public Public domain/
NDMA, SOI/Census/LG ASTER Physical survey/
domain Census 2011
SDMA,BMTPC Directory public domain Public domain

Superimpose
the layers using
GIS Application
Geodatabase (ESRI)

District hazard
risk map

Source: Illustration provided by Senthilkumaran Krishnan and the Reliance Foundation disaster management team.

The first step of the programme was conducting mapping was a key part of the needs assessment.
a sector-specific needs assessment between 31 Respondents reported their assessments of the
October and 4 November 2022 through focus group likelihood of flood, river erosion, earthquakes,
discussions. Ninety-one participants, including drought-like situations, hailstorm, and Nor’westers
people in vulnerable occupations and government occurring in each month of the year, and the
stakeholders from three Cachar gram panchayats, a
impact this would have on infrastructure
attended the four discussions. Hazard seasonality and livelihoods.

a
The gram panchayat is a village-level governing body under India’s Panchayati Raj system of rural local self-government.

50
Deshabandhu Club, which is one of the Cachar plan accurately, each space’s existing amenities
DDMA’s network NGOs for disaster management, also (including sanitation facilities) and its holding
attended the discussions. Other attendees included capacity (the number of people that can be
members of the Inter Agency Group (IAG)b Cachar, accommodated during a disaster) need to be
and representatives from the Community Facilitation assessed. The Reliance Foundation team conducted
and Resource Centres (CFRCs). critical infrastructure mapping and geotagged
these public sites. Critical infrastructure mapping
The CFRCs began in October 2022 as a joint initiative would ideally support local governments in planning.
of the DDMA and panchayat and rural development For instance, Panchayati Raj institution members
departments in collaboration with the IAG to build or community planners will be able to identify any
disaster resilience in the community. Ultimately, existing gaps and make a development plan to add
CFRCs are intended to become the repositories of facilities to enhance community-level preparedness
data and information, able to give advice during and and crisis management.
after disasters to officials regarding the panchayat,
and to the families residing within it. For example, in Balichara Grant Pt - II village in the
Borkhola panchayat of Cachar’s Silchar taluk, geo-
In November, the needs assessment was presented to coordinates of 33 vulnerable households and public
the Cachar DDMA, which then shared these findings utilities such as a school, health centres, roads,
with the heads of departments working at the community halls, and water canals were mapped.
district level for disaster management (such as the
departments of social welfare, state rural livelihood Reliance Foundation used secondary data and
mission, and agriculture), and grassroots NGOs. historical information to identify which parts of the
A crucial conclusion was that the community that district were at risk of flooding, marking villages
would be affected by disasters did not have adequate as being at ‘very high,’ ‘high’ and ‘moderate’ risk,
information on the basic facilities available in the and created a flood layer map with GIS. Then, using
vicinity for protection and shelter, while the government a digital elevation model from ASTER (Advanced
departments lacked updated information. Reliance Spaceborne Thermal Emission and Reflection
Foundation also demonstrated the capabilities of Radiometer), a flood inundation map was created.
critical infrastructure mapping through geotagging Villages that came under the ‘very high’ and
(the process of adding geographical metadata, most ‘high’ risk categories are the most vulnerable, with
often geographic coordinates, to a map). 209 such villages being identified so far.

Public facilities like health centres, hospitals, schools, These geo-coordinates were superimposed on a
anganawadis c
and panchayat offices are crucial flood inundation map and the map was presented
during disasters, serving as shelters for vulnerable to the DDMA, to demonstrate the potential of
people who need to be evacuated. To be able to geotagging and GIS as a tool for their use.

b
A group of organisations that includes civil society bodies, government representatives, representatives from UN
organisations, and others who work towards humanitarian activities in a particular region.
c
Anganwadis are a type of rural childcare centre. They were started by the Indian government in 1975 as part of the
Integrated Child Development Services programme to combat child hunger and malnutrition.

51
Image 2: Process diagram for Balichara Grant Pt – II village

ASTER Digital
Elevation Model Village map with
(Low lying area administrative
-extracting the boundaries, hydrological
elevation of the Most vulnerable villages under high and very features, household,
water surface from high flood hazard critical infrastructure etc.
a DEM)

Historical Flood Map

Source: Bhuvan Data
Products - Indian
Space Research
Organisation

Historical Flood
Hazard Map Source:
National Remote
Sensing Centre
Graphic representation
of earth surface
with administrative
boundaries, household,
critical infrastructure
mapping

GIS-based digital model

Source: Illustration provided by Senthilkumaran Krishnan and the Reliance Foundation disaster management team.

Similar maps can be created for cyclone risk or other Once there is data identifying the vulnerable
hazards. Reliance Foundation has already begun households, evacuation plans can be prepared on
creating multi-hazard maps, taking into account maps. By identifying how many persons are present
that facing multiple hazards increases an area’s in the household that would need to be evacuated,
vulnerability. mapping out a safe spot for them, and the path to
reach such shelters, the residents and officials can
The CFRCs had begun household surveys since their be prepared for when the next disaster strikes. The
inception. The process of collecting household level information can also be used to assist people in
data became easier and was made more secure, with accessing social entitlements, such as a disability or
Reliance Foundation assisting the Cachar DDMA in old-age pension and or ration cards, reducing their
preparing surveys using mobile phones. Detailed data vulnerability.
on every household is collected to identify indicators
that make them vulnerable in the event of a disaster. The DDMA retains the raw data and intends to share
The data points include the number of members; their only anonymised data with appropriate partners,
age, gender, education levels, and occupation; if there including Reliance Foundation, for the purpose of data
are any persons with disabilities; if the houses have analysis and planning for interventions. They were
temporary and unsafe roofs; and if the household initially cautious regarding the collection of data, as
has access to social security and has government some of the data is sensitive, but the development of
identification cards. Geotagging is automated, with the survey and data collection process was reassuring
individuals filling in the data via a phone. since no one but the government can access the raw
data.

52
IMPACTS AND OUTCOMES

Local volunteers and stakeholders, including the to over 61,000 individuals during Cyclone Sitrang
Border Security Force, were key in supporting relief (October 2022).
operations in Cachar during the 2022 floods. Manar
Shylla, a volunteer, said, “It was humbling for me to It is well-recognised that disaster risks are
work for the Tribal Khasi community who are one of localised, and responses must be so as well. In
the most vulnerable community.” Understanding the 2022, the Ministry of Panchayati Raj developed
various aspects that lead to vulnerability, including the Disaster Management Plan with the aim of
socioeconomic, is vital to reduce risks and is a first developing disaster resilience at the grassroots level
step towards building resilience. among the panchayats. A key aim was to ensure
a participatory planning process, and community-
Reliance Foundation has conducted two training based disaster management. The ongoing work in
sessions, one with local NGOs (including the IAG, Cachar is an example of how partnerships between
and the Deshbandhu club) and the other with local governments and all stakeholders can drive
members of the Cachar DDMA. The training focused progress in disaster management.4 Cachar district’s
on the collection of primary data—how to use the existing data on infrastructure and GIS based
household surveys that have been prepared, how maps could not be utilised systematically. With
to accurately collect the data, and how to use the detailed assessments of household-level of
geotagging feature—enabling them to conduct the vulnerability and infrastructure, and superimposition
data collection activity. Subrata Das, community of appropriate geographical features, the district
volunteer in Cachar’s Katigorah, said, “For our will be able to plan better and the community will
community resilience building planning, we needed be better prepared.
information on the details of vulnerable households
and its members. This information would help us As the impacts of climate change continue to
in providing social entitlements.” be increasingly felt, humankind must prepare for
more extreme conditions and worsening disasters.
While the Cachar DDMA retains the collected data, Already vulnerable communities can no longer
it has partnered with nodal non-government rely on their traditional knowledge of disaster risks
agencies such as Deshbandhu Club to mobilise the and protection. Systematising the process of
community. Reliance Foundation is also assisting assessing vulnerability, providing knowledge to
the organisation in some areas, such as initiating the local governments, and planning to make and
‘lightning safety’ advisories and livelihood advisories process quick decisions during disasters is vital.

53
KEY LESSONS
• Partnership with multiple stakeholders, including local governments and humanitarian organisations, can
create a multiplier effect. It aids in gaining the trust of the communities and in using localised knowledge,
and the network allows for greater outreach. Additionally, it is expected that aiding local government
stakeholders in planning will create a stronger and more sustainable model than working alone.
• Disaster management requires localised solutions, with the support of local stakeholders who are aware of
the needs and concerns of the region. Disaster management at the state level needs to be supported further
at the local levels.
• Geospatial data combined with data on vulnerability can be used to enhance community preparedness.
ENDNOTES

1
S, Vidya, and Indu K Murthy. “District-Level Changes in Climate: Historical Climate and Climate Change Projections
for the North-Eastern States of India.” Center for Study of Science, Technology and Policy, June 2022. https://
www.cstep.in/drupal/sites/default/files/2022-06/North%20Eastern%20Region_Final_10.06.22%20(1).pdf
2
“Assam Floods: Over 55 Lakh People Affected across 32 Districts, Toll Rises to 89,” The Hindu, June 22, 2022,
https://www.thehindu.com/news/national/other-states/assam-floods-over-55-lakh-people-affected-across-32-
districts-toll-rises-to-89/article65553525.ece
3
T J Cova, “GIS in Emergency Management,” in New Developments in Geographical Information Systems: Principles,
Techniques, Management and Applications, ed. Paul A Longley et al. (Wiley, n.d.), pp. 845-858.
4
Ministry of Panchayati Raj, Government of India, https://pib.gov.in/PressReleaseIframePage.aspx?PRID=1807011.

55
56
‘MINDSPARK’:
WORKING TO DELIVER
EDTECH FOR ALL

Antara Sengupta

T he COVID-19 pandemic adversely affected one of

the world’s largest schooling systems, with millions
Additionally, Lucknow was chosen as the area of
intervention as, in 2019, none of the government
of children being left out of school when India went schools within the intervention sample had any
into lockdown in March 2020. Although e-learning digital infrastructure nor were these being provided
solutions were already being experimented with by other NGOs. Under this intervention, learning
pre-pandemic, it emerged as a potential solution to labs were established in 55 government schools
tackle the learning losses induced by the pandemic. in urban and semi-urban areas in the city. Students
Unfortunately, most resource-poor schools could not from Class 1 to 8 were exposed to Mindspark with
access education technology (EdTech) solutions and around 30 hours of learning per year per subject.
adapt to the change, resulting in further deterioration It also involved providing support to teachers on
of learning levels among children in the country. platform management for long-term sustainability.
To be able to accurately understand the change
In India, even before the pandemic, only about 44 in learning outcomes, this intervention needed to
percent of children in government schools in Class rely on data throughout (from the design stage to
5 could read a Class 2 level text. Indeed, “The pool the assessment stage) so that the programme
of children that government schools were drawing could be improved at intervals to help children
their students from…became steadily more learn effectively.
disadvantaged.” 1
In order to enhance learning levels
and make EdTech accessible to children from low- The EI-PIF intervention is part of the Quality
income communities, Pratham InfoTech Foundation Education India Development Impact Bond (QEI
(PIF) and Educational Initiatives (EI) launched a DIB), led by a consortium of partners, including
computer-based adaptive learning programme the British Asian Trust, Michael & Susan Dell
in 2019 in government schools in Lucknow, Uttar Foundation, and UBS Optimus Foundation. Impact
Pradesh (UP). EI and PIF collaborated with the UP bonds essentially places a focus on outcomes
government to run a pilot of the initiative in the rather than outputs. In addition, these programmes
state to address the low enrolment rate (around transfer risk from government (as the payer of results)
43 percent) and learning outcomes in government
2
to the NGOs delivering the results. As these NGOs
schools. UP is India’s most populous state, with often cannot bear such financial risk, they may choose
around four crore school-going children (0-8 to engage with private sector investors to assume
years), which is 19 percent of the country’s total this risk. The resulting structure creates strong
school-going population (as of 2018). Lucknow incentives for the education partner and for investors
is UP’s most populous city, and as such is a good (if these are engaged) to deliver results. This leads
representative for potential scale-up and replicability to an improved procurement process and better
in the rest of the country. value for money for government spending.

57
LEVERAGING DATA FOR DEVELOPMENT

The accurate measurement of data and results are at home was not a useful indicator to gauge if
the cornerstone of outcomes-based financing. The children can optimally use it to learn effectively.
Mindspark programme began with clear outcome “Hence, we provided up to 25 Chromebooks to
indicators (enrolment and learning gains) specified in each school so that students can operate the
the evaluation design. The specific outcome targets device themselves and learn on them in some
were discussed with individual implementation dedicated periods within their classroom hours,”
partners yearly, which helped them register a baseline said Yadav.
and continuously track progress.
They were also aware that teachers needed
At the ideation stage, EI-PIF teams studied the continuous support and training in the software as
Annual Status of Education Report (ASER), a yearly well as platform management to effectively deliver
survey by Pratham that provides data on the status the lessons. This was also pertinent to ensure the
of schooling for children and their learning levels sustainability of the initiative.
in rural areas in India. “ASER data indicated the need
for an EdTech intervention in this region but we “We employed sancharaks (communicators) for each
wanted to design something that would not only school, who would help run the programme and
match the existing learning levels of the students, assist teachers,” Yadav explained. “They are local
but also help teachers track and improve their youth aged between 25-30 years, who are trained
performance consistently,” said PIF co-founder and and skilled under this programme on how to use the
CEO Prem Yadav. The Mindspark intervention aimed digital medium to use EdTech.” However, sancharaks
to achieve at least 10-15 percentile of grade- cannot be replaced by teachers as they do not have
appropriate learning levels for the target group. subject knowledge.
Using ASER, the team had insights into UP’s status
on enrolment and attendance patterns, learning In the first year, the team conducted internal
levels, and school facilities. assessments to understand each child’s learning
levels for Math and Language. Children were grouped
Besides the ASER data, they also conducted based on these assessments to facilitate peer
recces and surveys in about 110 schools in Lucknow learning. After each lesson, tests were generated
(from a list provided by the state government) to based on students’ learning levels, which helped the
understand the baseline data. The survey helped them EI-PIF team track each child’s progress.
understand that the availability of smartphones

58
Figure 1: An assessment page on the Mindspark interface

Implementation

Under the QEI DIB, EI-PIF developed clear students each day, which was collected and stored
assessment metrics to improve the performance in a central database.
tracking of schools and students. This helped in
early problem detection and deploying targeted Based on the data, the five best and worst
solutions that were informed by data. For instance, performing schools were identified. The schools
EI-PIF tracked the daily usage of Mindspark with the best outcomes were motivated to
software as well as indicators on the quality of share best practices, whereas an executive
usage (the number of consecutive questions committee at the programme level deliberated on
answered correctly by students). This directly solutions to mitigate challenges encountered by
impacted student learning outcomes and informed the five worst schools in terms of achieving
EI-PIF’s actions in schools or students that were learning outcomes. The EI team also
lagging. Notably, “Whenever a child solved three conducted workshops to train the PIF team on
questions correctly, the student would get a ‘spark’ understanding and reading the system-generated
on the screen. This motivated them and seeing data. “The EI team created functions as per our
their excitement the others were also motivated,” requirements and the data generated comprised
said Roopa Billava, Programme Lead at PIF. both quantitative and qualitative data. These were
simple data points, which were broken down for
Students were given log-in credentials to access us to understand the gaps,” said Billava.
the lessons and the Mindspark software on the
Chromebooks, which was connected to a server The teachers were then further trained and
and would be synced before and after the class capacity-building exercises were designed to help
each day. All 55 labs generated data on them find solutions to the problem. “In a physical

59
classroom, teachers cannot pay complete attention Challenges and solutions
to all students and cannot possibly know the
weaknesses of each child,” Yadav said. “As they Given the nature of technology, the remote
were trained in platform management, they could location, and strength of the internet, the
also understand the challenges that students face, team sometimes faced challenges in system
including which topic they are lagging behind.” synchronisation, which delayed the process of
These teachers would then take extra classes to data generation.
help the students overcome their gap areas.
Often, children faced difficulties in reading the
In the first year of operations, the team had set a questions on the screens. To address this, the EI
target of completing 20-hours of EdTech usage for team created audio clips, so they could listen to
each child annually. “However, we soon understood the questions and answer it by typing it out or
that the teachers find it difficult to understand selecting the correct option on screen.
the breakup and we were not able to achieve the
minimum usage hours,” said Yadav. After analysing However, the team encountered the most
the user generated data, the team launched significant challenge when the country went into
‘Mission 35’, encouraging 35 minutes of computer lockdown due to the COVID-19 pandemic and
use per day per child. schools were shut.

The level of data access differs across “We went to each stakeholder, including the
stakeholders—teachers and school management government officials, management, principals,
had access to school-level granular data; the parents and panchayats a
to understand how we
district or block offices had access to can reach the students,” Billava said. “It was
aggregate data for all schools within the block heartening to see that all of them were keen on
or district; sancharaks had school-level data; and continuing education.” The community offered
the EI-PIF team have access to aggregate data space to the EI-PIF teams to conduct
and granular data for all students in each school community classes.
across locations. Each stakeholder tracked
and took ownership for their part of data and EI-PIF launched an app version of Mindspark that
helped students achieve grade-level learning could be accessed by students with smartphones
outcomes. at home. In addition, they delivered SMS-based
lessons for parents to conduct learning activities at
As for data privacy, the EI team explained that home. They tracked weekly usage and attendance
the data collection is non-invasive and based on data, informed teachers about questions that
an algorithm. “It is in situ data capturing, hence were routinely answered incorrectly and built SMS-
there is no need to ask for permissions,” said EI based lessons on such questions.
Vice President Ritesh Agarwal. “We do not share
any individual data publicly and only aggregated
data is shared with the programme stakeholders.”

a
Village-level governing bodies in India.

60
The intervention transitioned from an in-school They taught children, and carried the laptops back
model to community classes, conducted in homes, in the evening, only to return the next day. The
temples, or community halls. community members also helped us a lot. In an hour,
10 children would learn with the volunteers on
Om Prakash Yadav, Operations Lead at PIF, Mindspark. And so, over six hours each day, 60
recalled, “The volunteers worked a lot. They children would be taught. We tried our best to
charged the laptops in their homes if needed reduce any losses in education for the children.”
and carried extension boxes to the community.

IMPACTS AND OUTCOMES

The team adopted a mission-mode culture and By the end of the QEI DIB, in 2022, students
used data extensively to improve performance on under the EI-PIF intervention recorded a near
outcomes. This allowed them to exceed learning and fivefold improvement in literacy and numeracy skills
enrolment targets each year. in comparison to non-participating students
(See Figure 2).

Figure 2: Average learning gain per year over four years

Learning outcomes are measured using

PinAcLe (Progression in Achievement
of Learning), which is the assessor’s
vertically integrated learning level scale
where scores from Grade 1 to 10 range
from 200 to 800 scale score points.

Control group Treatment Group

61
“The computers were an entirely new thing for them. “The aggregated data was shared even with the local
They learned language and mathematics using administration, which helped them conduct regular
them, in classes that lasted about 45 minutes. From meetings and workshops to address critical gaps,”
the books they used to solve about 15 questions said Yadav. The data empowered each stakeholder in
before they got bored and distracted. Now they are the ecosystem and enhanced the overall performance
completing around 50 in each session,” said Yogendra levels.
Kumar of PIF.
PIF has now taken this learning forward and instead
Although students were the direct beneficiaries of of recording monthly attendance manually, all their
the intervention, teachers and parents too benefitted other school programmes track daily attendance
through various awareness and capacity-building through technology.
exercises. For example, during COVID-19, EI-PIF teams
conducted a cybersecurity awareness campaign for “It is most important to build a culture of data-driven
parents since students were using the mobile app decision-making in the programme ecosystems,”
version of the Mindspark software. Teachers were also said Yadav. “The idea of enticing children with sparks
trained in platform management and in running the is to ignite enthusiasm in children to perform better.
software. Since teachers have been trained to handle Similarly, categorising schools as best and worst also
the operations on their own, the number of sancharaks helps them enhance or improve performance, with
has now reduced by half. respect to other schools.”

“We have been able to affect behaviour change in While the EI-PIF programme concluded in March
the intervention schools for using edtech and now 2022, PIF continues to support the 55 schools.
even control schools are asking us to implement the Importantly, given the programme’s exemplary
programme in their schools,” said Billava. results, NITI Aayog is now replicating the programme
in other parts of UP.
Besides, the ‘division of labour’ for data collection
has created marked efficiencies in implementation.

62
KEY LESSONS
• Real-time data generation helps generate micro insights and facilitates quick and timely decision-making.
• It is important to build capacities of each stakeholder in understanding data points. This helps in faster
collective action and reduces any time lags in decision-making.
• An outcome-based financing mechanism is a strong driver for better educational performance, but it relies
on data and accurate measurement. Edtech solutions, by their nature, are well suited to be integrated into
such a financing structure and can drive better learning outcomes.
ENDNOTES

1
“Annual Status of Education Report 2018”, ASER Centre, accessed February 6, 2023, https://img.asercentre.org/
docs/ASER%202018/Release%20Material/aserreport2018.pdf
2
“Annual Status of Education Report 2023”, ASER Centre, accessed February 6, 2023 http://img.asercentre.org/
galleries/aser2021_final.pdf
3
“Handbook 2018”, Statistics on Children in India, National Institute of Public Cooperation and Child Development
https://www.nipccd.nic.in/file/reports/handbk18.pdf

64
THE ‘PATH’ TO USHERING A
NEW ERA OF DIGITAL HEALTH
Noyontara Gupta

In 2021, eight countries accounted for more than

two thirds of global tuberculosis (TB) cases. About 1
According to the National Sample Survey Office, as
of 2014-15, over 70 percent of healthcare services
28 percent of these cases were concentrated in India were provided by the private sector. About
in India, with Indonesia, China, the Philippines, 72 percent of rural residents and 79 percent of urban
Pakistan, Nigeria, Bangladesh, and the Democratic residents used private healthcare services.3 Moreover,
Republic of Congo also having a high incidence nearly 81 percent of India’s doctors work in the
of TB. private sphere. The diagnosis of TB and treatment
status of patients in the private sector was neither
TB has persistently been among the most persistent recorded nor notified in government information
public health challenges in India. Even so, the systems, inevitably leading to underreporting,
country has committed to eradicating TB by 2025, which in turn led to the suboptimal monitoring of
ahead of the target set under the 2030 Agenda for treatment completion rates. Many of these challenges
Sustainable Development. have persisted, and close analysis of patient care
in private-sector TB treatment has revealed more
In its efforts to tackle the disease, the Indian concerns, such as delayed diagnosis, less-than-
government has undertaken several measures— optimal quality of care, and financial burdens
implementing the Revised National TB Control due to high out-of-pocket expenses, leading to a
Programme, providing free TB diagnostics and high rate of treatment drop-out. In a country like
drugs to patients, and making it mandatory for India, where public health facilities are overburdened,
practitioners to notify TB cases through the for health initiatives to see true success, have
government’s web-enabled patient management large-scale impact, and be wholly inclusive, acceptance
system for TB (Nikshay). The government also in private healthcare is crucial.
developed the National Strategic Plan 2017-2025, 2

which outlined various approaches to eliminate

TB by 2025. One key strategy noted in the plan was
private-sector engagement.

LEVERAGING DATA FOR DEVELOPMENT

PATH, an international non-profit founded in 1977, a pilot to address this gap in high slum population
involves a global team of innovators to accelerate wards in Mumbai was eventually scaled up across the
health equity. The organisation identified a lapse in country, firmly establishing PATH as a leading force
TB response across the country, especially among in private-sector engagement efforts in India’s fight
private-sector healthcare providers. What began as against TB.

67
The story begins in 2014 with PATH’s Private delivery model was adopted, wherein incentives were
Provider Interface Agency (PPIA) programme, which offered to patients in the form of vouchers (to get
was launched as one of the initiatives under the access to free World Health Organization-approved
larger umbrella of ‘Mumbai Mission for TB Control’ diagnostic tests and medicines from PPIA-networked
under the aegis of the Municipal Corporation of laboratories and chemists) and for treatment
Greater Mumbai. Through PPIA, PATH reached support for patients to complete their treatments.
out to private healthcare providers to enable Patient support services were led by Maharashtra
early diagnosis, prompt response, and timely and Janvikas Kendra and ALERT-India, two organisations
increased notification of TB patients, enabling that have been working in Mumbai slums for over
patients to see through the entirety of their three decades. By suitably leveraging technology,
treatment. Identifying the key role played by patients were also sent daily reminders via SMS
disruptive innovations like technology and by and through dedicated call centres to ensure
leveraging market forces, PATH actively worked treatment adherence.
towards developing a collaborative relationship
with the private sector to facilitate TB control The data collection process was not easy. The
in Mumbai. The PPIA model utilised digital project involved the collection, storage, and
health, management principles, data, research, monitoring of data that was related to private
information and communication technologies (ICTs), providers and TB patients. With over 13,000
and frequent stakeholder inputs to work towards medical facilities in the Greater Mumbai
achieving its goals. The model was designed to area in 2014, manual databases were not feasible.
enable TB patients in the private sector get access Further challenges emerged while mapping
to free public-sector drugs and subsidised state-of- geographies and narrowing down data that could
the-art TB diagnosis. be relevant. In 2015, a platform was developed
that could host the end-to-end patient
According to PATH representative Rishabh data—beginning with the notice of a potential
Chopra, the objectives were simple: increasing the TB diagnosis and ending with the outcome(s)
number of TB notifications from private healthcare of the treatment. The platform, named Universal
providers; ensuring that providers maintain Access to TB Care (UATBC), offered services like
appropriate quality of healthcare as per the notifying confirmed TB cases to Nikshay,
Standards of TB Care in India; tracking and ensuring generating electronic vouchers (replacing the
treatment adherence (to see if the patient actually paper voucher incentives that were distributed earlier),
went through with treatment post-diagnosis), and providing CB-NAAT testsa and first-line anti-TB
analysing the post-treatment outcomes. drugs,b reminding patients to follow up on
treatments, and checking in on patients
PATH deployed field officers, who were often former periodically to monitor treatment. The UATBC also
pharmaceutical representatives, led by personnel helped provide digital identification numbers to
in the pharmaceutical market, to approach private- the providers, and facilitated the generation
sector healthcare providers and patients. A service of e-prescriptions.

a
Widely accepted diagnostic test for TB.
b
First-line drugs are first choice drugs for treating a condition with the least likelihood of causing side effects. Of the
approved drugs, isoniazid (INH), rifampin (RIF), ethambutol (EMB), and pyrazinamide (PZA) are considered first-line
anti-TB drugs and form the core of standard treatment regimens.

68
The PPIA TB pilot programme, which ran from Further enabling the case notification and treatment
2014 to 2017, saw the successful engagement of patients, the government has incentivised the
over 13,212 private providers (which included process through the Nikshay Poshan Yojana, a
practitioners, specialists, chemists, labs, and direct benefit transfer scheme for TB patients.5 Under
more). The numbers are also impressive in terms of this scheme, patients receive INR 500 per month
patient coverage—about 1,80,000 people availed for the treatment duration of six months. The
PPIA services, 60,366 TB patients were notified, government has also made concerted efforts
and nearly 80 percent of those who underwent to engage providers in case notifications, offering
treatment completed the course successfully. INR 100 for every TB case that they notify. These
efforts have gone a long way towards increasing the
The effective use of data and technology for visibility of TB cases and treatment adherence
development is abundantly clear. For example, for patients.
according to PATH, since it was possible to
capture events in a patient’s course of treatment, In 2022, based on the learnings from the PPIA,
once aggregated, such data would provide PPSA, and JEET (2014-2021), PATH launched a
population-level insights into treatment adherence. pilot in Mumbai to support the National Health
The model also enabled the quicker notification of Authority to onboard private sector healthcare
cases and transfer of data from the private sector into providers onto the Ayushman Bharat Digital Mission
the public sector health systems, leading to the (ABDM), a task that includes the registration of
inclusion and greater adoption of ICT tools in health professionals and health facilities (clinics,
government notification systems across the country. nursing homes, hospitals) on relevant ABDM
registries, the creation of Ayushman Bharat Health
Three years into the pilot programme, with the Accounts (ABHA) for patients, generation of
support of the Indian government, a pan-India electronic health records, engaging with stakeholders,
scale-up of the PPIA model was rolled out, finding a and more. The pilot aims to explore new ideas on how
new identity as the Patient Provider Support to increase the adoption of digital tools in
Agency (PPSA). healthcare and improve access to healthcare for
patients by utilising technology.
The evolution from PPIA to PPSA served as the basis
for the Joint Effort for Elimination of TB (JEET) Like the PPIA and JEET, the new pilot is built on the
project. Jointly implemented by the William J. Clinton premise of leveraging technology to strengthen and
Foundation, Centre for Health Research & Innovation develop India’s health ecosystem. As of January 2023,
(a PATH affiliate), and the Foundation for Innovative the pilot has onboarded over 300 health professionals
New Diagnostics, with funding from the Global and facilities, resulting in the generation of 3,000
Fund to Fight AIDS, Tuberculosis and Malaria, Project ABHA-linked electronic health records.
JEET upheld one main mantra, “go where the patients
go,” says Chopra. Notably, the goal of the ongoing pilot is not just
to onboard healthcare professionals onto the
By scaling up the programme, India’s ‘trace, track, ABDM, but to truly digitise their practice and
and treat’ strategy to combat TB has been informed develop a model for engagement and onboarding of
by the data and insights that emerged out of PPIA, private-sector healthcare providers on the platform.
and subsequently the PPSA. Project JEET has been a Moreover, through ABHA-linked e-health records,
significant contributor to total notifications from the the patient will have access to longitudinal health
private sector of TB cases, increasing from 29 percent records that can be easily shared with healthcare
in 2018 to 68 percent in 2020.4 providers, creating a feedback loop and transparency
that would help inform diagnosis, treatment,
and follow-ups.

69
By leveraging the pre-existing networks created The pilot has been a massive undertaking—bringing
during PPIA, and with support from the Municipal in local ward-level medical officers and associations
Corporation of Greater Mumbai and the state to engage doctors on the need to digitise their
public health department, a list of doctors has been practice by means of in-person sessions, interactive
mapped across selected wards in Mumbai. The videos, flipcharts, posters, and pamphlets. The
list is not exhaustive, but gives a broad idea of the process has not been without challenges. Extensive
contours that have to be covered. PATH deployed training has been provided to the field officers,
a field interface agency to reach out to the clinics, such that they could approach and inform
identifying the number of doctors in the process and the clinics and doctors accordingly. Moreover,
onboarding them to ABDM. constant feedback is being obtained on the user
experience for continuous improvement and
According to Tanu Gupta from PATH, an internal evolution of digital solutions.
dashboard has been created to maintain records
and monitor progress, which the field staff update Under the new pilot, PATH is working towards the
on a daily basis. The dashboard maps various targeted sensitisation of doctors to adopt digital
data points—from progress updates to notes on solutions by telling them about ABDM, immediate
which doctors are more receptive and which are and long-term benefits of digitisation, and addressing
more reactionary. The database offered significant their inhibitions and concerns by giving further details
insights on whether the pilot was working, and what about the mission and citing examples from Indian
could be done to make it more effective. digital success stories (such as Aadhaar, Unified
Payments Interface, and CoWINc).

IMPACTS AND OUTCOMES

While the Mumbai pilot is still at a relatively early driving a multi-partner initiative and collaboration
stage, it has begun to demonstrate impressive for impact. Digital tools can enable data-informed
outcomes. More and more benefits will become corrective actions that help improve service
visible once the Unified Health Interface delivery and reinforce the confidence of the end
becomes fully operational. The project has already
d
users. This supports improvements in the quality
begun to stand out as a critical step towards of care while simultaneously easing the burden on
transforming the digital healthcare landscape in an already-strained ecosystem by bridging the
India, particularly in light of the COVID-19 pandemic. divide between public and private providers. Through
these efforts, PATH is paving the way for change.
Stakeholders in healthcare provision cannot work
in silos. PATH’s efforts have shown the value of

c
Unified Payments Interface is a mobile payment method, wherein your smartphone can essentially function as a debit
card. CoWIN is the Indian government’s web portal for COVID-19 vaccination registration.
d
The Unified Health Interface is a network of open protocols that enable the interoperability in health services.

70
KEY LESSONS
• A system of periodic assessments ought to be established at the state and national levels to identify
gaps in public health delivery. Based on their findings, strategies ought to be evolved for engaging and
incentivising private healthcare providers to address these gaps.
• Civil society organisations could play a critical role as intermediaries connecting private service providers
to large-scale government initiatives, or helping them build applications or tools based on existing digital
public infrastructure.
• The potential for scalability and replicability must be integrated into the basic design of digital health
pilots or proof-of-concept projects. If found to generate value, such initiatives often need to scale-up
rapidly across geographies.
• Dashboards can be indispensable to the management of ‘data for development’ initiatives. By presenting
real-time data and trends to decision-makers, they enable policymaking, course correction, and strengthen
transparency and efficiency.
ENDNOTES

1
Global Tuberculosis Report 2022, World Health Organisation, https://www.who.int/teams/global-tuberculosis-
programme/tb-reports/global-tuberculosis-report-2022/tb-disease-burden/2-1-tb-incidence#:~:text=In%20
2021%2C%20eight%20countries%20accounted,2).
2
National Strategic Plan for Tuberculosis: 2017-25 Elimination by 2025, Ministry of Health and Family Welfare, 2017,
https://tbcindia.gov.in/WriteReadData/National%20Strategic%20Plan%202017-25.pdf
3
Nishant Jain et al, “NSSO 71st Round: Same Data, Multiple Interpretations”, Economic & Political Weekly, November
21, 2015, https://www.niti.gov.in/sites/default/files/2019-01/NSSO_71st_Round%20_Final.pdf
4
Joint Effort for Elimination of Tuberculosis (JEET) 2018-20, https://www.projectjeet.in/wp-content/
uploads/2022/01/JEET-report-2018-2020.pdf
5
Chaitanya Deshpande, “TB notifications up 23%, but it’s good news for state”, The Times of India,
July 17, 2019, https://timesofindia.indiatimes.com/city/nagpur/tb-notifications-up-23-but-its-good-news-for-
state/articleshow/70250820.cms

72
‘CLOTH FOR WORK’:
USING DATA TO BUILD
A CIRCULAR ECONOMY

Swati Prabhu

C ommunity
bedrock of
development
good governance
is considered
and
the
sustainable
smooth supply chain management, particularly at
the grassroots level. This can be a disaster-stricken
societies. A powerful and effective tool, it can area or a remote village. Demographically, their
potentially help tackle a host of social challenges, efforts cut across barriers of caste, age, religion,
ranging from poverty to livelihood generation.
1
and gender. Goonj’s activities centre around the
However, the Indian social sector faces several ‘Cloth for Work’ model—a bottom-up community-
challenges while fulfilling this goal.2 One critical led development approach underlining
problem is in ensuring the successful delivery of the importance of material as a tool of social
support and effective logistics management at development. Here, the central objective
the grassroots level in emergency situations, such is to utilise urban surplus material through
as a pandemic or disaster event. Indeed, disasters community sensitisation, which acts as a motivator
can wipe out years of development by destroying to enable large-scale development activities on
habitat, crops and infrastructure, resulting in vast the ground. It collaboratively devises solutions
human displacement.3 In such a context, a circular with communities to address issues as diverse
economy may prove particularly effective. Offering as menstrual health and hygiene, education,
a systems solution framework, a circular economy employment, agriculture, water, and the like.
promotes the reuse of waste products by turning ‘Cloth for Work’ promotes the repositioning of
these into critical resources, in turn facilitating urban discard as a development resource.
resilient economies. 4
Through this model, Goonj has been collecting
and disbursing tonnes of material in 5000+
Goonj, an independent Delhi-based non-profit areas every year across 27 states and
organisation, attempts to use data to ensure union territories.

LEVERAGING DATA FOR DEVELOPMENT

The ‘Cloth for Work’ model relies on hybrid data engaging in freewheeling conversations and dialogues
collection techniques. The conventional mode of on pertinent issues of sanitation, health, local
conducting surveys (circulating questionnaires and infrastructure, water and community spaces, Goonj’s
conducting structured interviews) for gathering field teams help the local communities identify and
information helps, but Goonj’s advantage in data prioritise their material and social needs. In most
collection is its local grassroots workforce, which is cases, these team members are from the village
well acquainted with the cultural milieu, on-ground where they undertake extensive observation surveys
knowledge, and the rural world’s sensibilities. By to identify resource-deficient areas. This also has a

75
ripple effect on neighbouring areas, which office in Delhi, “Collecting data post disasters
often assess their specific needs, thus contributing to is a near impossible task because of the
the data-collection process. obvious bottlenecks in access, communication,
and the like in a real-world scenario. One learns,
The data collection process is different in urban observes, and understands the cultural practices
areas—it is more focused on sensitisation and and traditions of a specific region beforehand—
creating awareness about meaningful giving, about this is disaster preparedness. For instance,
avoiding mindless charity, and encouraging civic during the pandemic, no one had the time
participation to provide material to meet the to collect data and match it with their
basic needs of rural communities. Importantly, implementation strategies.” Now, the relief kits
data is collected to streamline the processing of made by the Goonj processing centres are
urban material to carefully match the needs of the carefully curated to match the needs of the
rural communities. communities on the ground by being mindful of
specific regional requirements.
Jigisha Maheta, a Goonj representative working
with the organisation’s Chennai chapter, notes Goonj’s data collection efforts involve gathering
how during the 2014 floods in Jammu and Kashmir, various sets of hard information, as well as soft data
the relief material collected from the urban centers in the form of experience and local knowledge. The
failed to match the dietary and clothing needs selection of partner grassroots organisations is
of those displaced. For instance, people from usually carried out by word-of-mouth and
Delhi and Mumbai sent basmati rice and saris, but background checks through information available
this did not meet the needs of those affected. in the public domain, such as media reports. This
In such situations, data collection and reconnaissance is essential from a functionality perspective. A
exercises can help sensitise the urban population judicious mix of both hard and soft data is crucial
about regional usage patterns, but this must be for the onboarding of partners. It also serves as a
done before a disaster strikes. According to Abhinav system of checks-and-balances to maintain the
Dutta, a Goonj representative from the head accountability and integrity of partners.

76
IMPACTS AND OUTCOMES

A primary challenge for Goonj is data storage. Given environment and have personalised methods of
that the organisation handles over 6,000 tonnes data collection. Villagers find it comfortable to
of material each year, it is tedious to maintain data share their opinions and experiences with the
related to this manually. Additionally, there is a risk Goonj team. This is also a result of the goodwill
of human error during the process of data entry. For generated by the organisation over
Goonj, the solution lies in going digital. According to the years. An awareness of the data
Dutta, “In the coming financial year (2023-24), Goonj collected in one village has a cascade
intends to digitise its data collection processes, thus effect on adjoining areas. People voluntarily
doing away with the overwhelming and complicated communicate their needs, thus facilitating the
paperwork. During the second wave of the COVID-19 process of project implementation.
pandemic, we took a giant leap and migrated to
Enterprise Resource Planning (ERP). Considering However, data collection assumed a different
the scale achieved during COVID-19, it has proved to meaning for Goonj during the COVID-19 pandemic.
be quite helpful in managing the material collection, Given the urgency of the situation, materials had
inventory, distribution, and reporting.” Dealing with to be dispatched and delivered quickly, with
more than 30,000 kgs of material every day, Goonj no scope to conduct recce and surveys. Here,
has realised that a user-friendly ERP tool facilitates Goonj’s familiarity with working in specific regions
the efficient and effective management of resources. came in handy. Still, another lesson learnt, was
that prior data collection and analysis may not
Over the past two decades, Goonj’s grassroots data always be useful in ascertaining the needs of a
collection networks have become deeply embedded community in a specific situation; for instance,
in the rural system. Grassroots organisations have during the pandemic, there was a greater need for
an excellent understanding of the rural cultural medical items rather than food.

77
KEY LESSONS
• Data collection at the grassroots level should be conducted independently, transparently, and
autonomously, with the minimal involvement of external agencies to avoid any dilution of information.
The lack of clear administrative data—such as on health, education, or nutrition—creates functional
difficulties for the social sector.
• The COVID-19 pandemic has shown that data collection can attain varied meanings and forms in different
disaster-like situations. Experience and local knowledge accumulated over the years will serve to
address the needs of the communities.
• Human-centric models of data collection can ensure the demands of communities are met in a positive
manner. This can be achieved by involving members of the local community.
• Logistics data collected from rural centres must be complemented by sensitising the urban
population to achieve a balance between demand and supply of the needed resources.
ENDNOTES

1
United Nations Economic and Social Commission for Asia and the Pacific (UNESCAP), “What is Good Governance?”,
https://www.unescap.org/sites/default/files/good-governance.pdf
2
Abhirup Bhunia, “India’s Social Sector has a Data Problem”, India Development Review, February 17, 2021, https://
idronline.org/indias-social-sector-has-a-data-problem/
3
Raheem A. Usman, F.B. Olorunfemi, G.P. Awotayo, A.M. Tunde and B.A. Usman, “Disaster Risk Management and
Social Impact Assessment: Understanding Preparedness, Response and Recovery in Community Projects” in
Environmental Change and Sustainability, eds., Steven Silvern and Stephen Young (London: IntechOpen, 2013)
4
Ellen Macarthur Foundation, “What is a Circular Economy?”, https://ellenmacarthurfoundation.org/topics/
circular-economy-introduction/overview

79
OUR APPROACH

M icro Matters is the third publication in a series

by Reliance Foundation and Observer Research
the eight exemplary case studies showcased in
Micro Matters.
Foundation about transformations in development
and governance. Through January and much of February, we
conducted video interviews with the programme
The idea for the publication began to germinate managers and project leads steering these eight
when India announced, in late 2022, that the interventions. The findings from each interview
principle of data for development (D4D) would be were supplemented with secondary documentation
integral to its upcoming term as G20 president. It provided by our CSO partners, and other information
struck us that while the D4D approach is not new gathered independently from government sources,
in itself, the evolution of new technologies is academic studies and media reports. We remain
allowing institutions to gather data and apply grateful to our partners for entrusting us with the
insights in ways that were not possible earlier. details of their initiatives.
We felt that presenting these models to a wider
audience could help export innovative D4D practices Each case study follows a similar format. We begin
not just to underserved locales across India but by introducing the context and aims of the D4D
also to other parts of the Global South. project in question. We go on to describe its
execution with particular reference to the methods
Accordingly, in January 2023, we began to enter of data collection and analysis used, the application
into dialogues with several India-based philanthropic of data-driven insights to specific challenges, and
and civil society organisations (CSOs) to better the various innovations employed. We conclude
understand some of their high-impact D4D by exploring the project’s impacts and outcomes
initiatives. These conversations led us to identify and deriving a few key lessons.

81
 ACKNOWLEDGEMENTS

A t the very outset, a special word

thanks to the philanthropic institutions and civil
of The team at Reliance Foundation, led by Jagannatha
Kumar and B Srinivasan, spearheaded this initiative
society organisations who shared their cases on to bring to the fore a diverse set of case studies
data-led development for this volume. Not only on data for development. Jayashree B and Vanita
is Micro Matters about their interventions, it is a Sharma were co-editors of the publication; Carla
guide to advancing development by collecting Gomez-Acebo Botin coordinated all operations;
data, processing it to evolve insights, translating Antara Sengupta, Ketaki Hate, Mohini Ganguly and
insights into actions, and making a social impact. Saji Kadavil contributed chapters; Sanjay Masand
and team, including Bonani Gupta Karmakar and
This publication would not have been possible Surya Kishore, co-led its design; and Deepa Tripathi,
without the time and support of several partner Taslim Arif and Shambhavi Chaturvedi supported the
organisations: the Alliance for Saving Mothers and production process.
Newborns (ASMAN), Anudip Foundation, British
Asian Trust, Digital Green, Educational Initiatives, The team at Observer Research Foundation,
Goonj, Michael and Susan Dell Foundation, PATH, led by Samir Saran, helped bring Micro Matters
Pratham InfoTech Foundation, UBS Optimus to life. Anirban Sarma and Shoba Suri were
Foundation, and Wadhwani AI. co-editors and led research and content
development; Noyontara Gupta, Shimona Mohan,
Our grateful thanks to: Anshu Gupta, J. P. Tripathi, Sitara Srinivas, and Swati Prabhu contributed
Monisha Banerjee, Nupur Bahl, Prem Yadav, Rhea chapters; Pratnashree Basu supported the writing
Miranda, Ritesh Agarwal, Ronali Pradhan, Sameer process; Preeti Lourdes John copyedited and
Kanwar, Senthilkumaran Krishnan and all their team proofread the publication; and Rahil Miya Shaikh
members for sharing their stories and insights co-led its design and production.
towards this publication.

82
T
INDIA 2023