Budownictwo i Architektura 14(2) (2015) 5-10

Dispatching concrete trucks using simulation method

Sławomir Biruk
Katedra Inżynierii Procesów Budowlanych, Wydział Budownictwa i Architektury,
Politechnika Lubelska, e–mail: [Link]@[Link]

Abstract: Ready mixed concrete (RMC) is the primary material required for build-
ings and public infrastructure work. RMC is produced to meet customer’s demands and its
deliveries must conform to construction site and technological operating constraints – the
material cannot be prepared in advance and stored. Concrete production scheduling and
truck dispatching is mainly handled manually by experienced RMC batching plants staff.
The paper presents simulation model which can be used to asses alternative strategies for
truck allocation and production planning in stochastic environment. The models’ operation
is illustrated by a notional case – the model prompted solutions of improved transhipment
efficiency and reduced plant operating cost under assumed operating constraints.
Keywords: ready mixed concrete distribution, truck dispatching, simulation model-
ling

1. Introduction
Ready mixed concrete is a basic construction material used in all components from
foundations to roofs, and therefore consumed in large quantities. As many as 60-70% of all
contemporary built structures are made of this material. Due to growing quality require-
ments and development in concrete mix design aimed at obtaining special qualities of the
product, it is manufactured by specialised batching plants in fully controlled environment.
The mix is delivered to construction sites by means of ready mixed concrete trucks assuring
that the mix keeps its qualities, but the time since mixing till placing is seriously limited:
even with retarders, concrete mix sets within several hours. Moreover, concrete mix should
be placed continuously to avoid cold joints that reduce strength and quality of concrete
members.
This poses a challenge to the mix dispatchers who need to coordinate mix production,
transport and placing processes. Most concrete batching plants dispose of advanced
automatic production lines able to provide high quality mix in a very short time, but the
dispatching process is still based on skills and knowledge of the experienced staff, with
little use of computer-based planning [11].
The ready mixed concrete production is a competitive business in a well developed
market. With similar material costs, the batching plants search for economies by means of
better dispatch control: as the material is relatively cheap and of relatively high transport
cost, an integrated system of mix manufacturing, scheduling deliveries and vehicle routing
may seriously improve the plant’s standing [3]. As the concrete mix expires quickly,
a batching plant may serve only those clients who are located in the range that can be
reached by trucks within about 1.5 hours. This fact seriously limits the number of potential
clients and intensifies competition between plants operating in overlapping areas. Thus, the
plants strive for low costs of their product and reliability of their service to stay in the
market [2].
6 Sławomir Biruk

2. Ready mixed concrete transportation process

The time since mixing till placing should be no longer than concrete setting time in
particular conditions. To deliver the mix within this “usability window”, one has to plan
vehicle routes and schedule production so that at the moment that the substrates have been
mixed, a truck was ready to collect the mix. Delays and breaks may lead to wasting the
whole batch. Having too many trucks queuing for the mix is also not economical.
The mix is a custom-made product: each batch is prepared for a particular client.
Therefore, as the batch is loaded into a truck, the truck must be directed to a particular
construction site. The dispatcher is in no position to change its destination. Moreover, is
quite usual that the construction sites served by a plant are congested or of limited
unloading capabilities and cannot receive more than one truck at a time. In this case, to
assure continuity of concreting works, the mix production and deliveries need to be
scheduled with utmost care.
Feng et al. [2] classify orders for concrete mix in two categories: those placed in ad-
vance – at least one day ahead of concrete works, or those placed at the day of concrete
works. The latter orders may seriously affect operations of a batching plant, and the
dispatcher needs to decide quickly whether accepting such orders is possible and how it
affects operating schedules.
The daily demand for the mix is not uniformly distributed. Feng et al. [2] observed
that peak hours occur between 9:30 and 11:30 and between 14:30 and 17:00, whereas
Matsatsinis [5] and Naso et al. [6] maintain that the demand is clearly higher between 7:00
and 9:00 and later between 13:00 and 15:00. The peak hours are especially challenging for
the production, dispatch and transport crews.

3. Optimisation models of concrete mix dispatching processes

The integrated system for production and dispatch scheduling is to allow for con-
straints specific for the concrete mix and fresh concrete. Time is the crucial factor, and this
makes mathematical model complex, and optimisation process difficult.
Tommelein and Li [9] approach this problem as a just-in-time production system.
They analyse the mix supply chains with regard to requirements of production process and
material qualities, putting forward two models of supply chains. The first one assumes that
the plant takes responsibility for the deliveries, the other one transfers all transport tasks to
the client (used sometimes for sites located in city centres with strict constraints of noise
emission or traffic disturbance – the works must be scheduled for particular hours).
Matsatsinis [5] developed a decision support system for managing two types of means
of transport: truck mounted mix pumps and ready mixed concrete trucks. This model was
formulated as a multi-depot multi-vehicle routing problem with time windows. The objective
function minimizes operating costs of the system and increases service quality level.
Feng et al. [2] investigated in detail into the factors affecting truck routing and sched-
uling. They proposed a model based on a genetic algorithm and simulation aimed at
minimizing truck queuing time. They provided also a user-friendly software facilitating the
process of dispatching concrete trucks.
A metaheuristic algorithm for the mix dispatching problem was proposed also Naso et
al. [6]. Their model is aimed at minimizing the cost of mix distribution. This cost includes
loading, transport, unloading, and additional costs related e.g. with hiring trucks and crew
overtime. The model was programmed in Matlab, and its applicability was illustrated by
a case of a Dutch batching plant.
 Dispatching concrete trucks using simulation method 7

Asbach et al. [1] put forward a mixed linear programming model of mix deliveries
with the objective function that minimizes the sum of transport costs and penalties for not
fulfilling the order. Due to the model’s complexity, it cannot be solved by means of popular
solvers, so the authors resorted to their own metaheuristic algorithm and assessed its
efficiency.
Park et. al. [7] expanded the scope of modelling from production and dispatch of the
mix by allowing for the batching plant supplies with cement and aggregates. The model
was solved by means of a customized continuous dynamic simulation system, and
a software facilitating operation of the plant was developed and tested in one of plants in
Singapore.
Yan et al. [11] integrated scheduling of mix production and vehicle routing. The
model is based on time-location diagram with five types of arches to represent operations of
the plant and the trucks (truck idle time, truck work, loading or waiting for loading,
unloading and returning to the plant, termination of work and returning to the base). The
model was then enhanced by allowing for additional constraints: crew overtime cost [13]
and traffic-related occurrences [12].
Yan et al. [10] considered that transport time is a random variable. Their dispatching
model was formulated as a mixed integer network flow problem with side constraints. The
model’s applicability was tested by simulations.
Computer simulations were frequently used in the analysis of concrete mix batching
plant operations. Wakefield and Sears [14] used Petri Nets, and Zayed and Halpin [15] –
a Microcyclone network. In both this cases the models were simulated with various
resource combination to evaluate productivity of a concrete plant and utilization of
equipment. Sobotka et al. [8] used simulation method to reengineer operations of a batching
plant. The aim of research was to improve production management by reducing activities
that, from the customer’s point of view, do not add value to the service. Lu and Lam [4]
developed a special-purpose simulation tool for rapid modelling of a typical one-plant-
multisite system of concrete production and delivery based on a simplified discrete-event
simulation approach. All these attempts speak in favour of applying simulation to assessing
and modelling operations of batching plants under a variety of conditions. However,
a general model that allows for all identified constraints in random environment.

4. The simulation model for dispatching concrete trucks

The author puts forward a simulation model for dispatching concrete trucks aimed at
assessing continuity of concreting works in constructions sites and the time since mixing
till placing concrete mix. It was programmed with GPSS World discreet simulation
language by Minuteman Software. The programme comprises three modules: declaration of
input, simulation engine, and collection of simulation data.
The first module requires that the following parameters are defined:
 number of destinations (construction sites) together with their demand for the mix
expressed by the number of truckloads to be delivered,
 as-planned start of concrete placing (hour) at each site, given as a delay from the
start of a work shift in the site,
 number of available concrete trucks,
 number of concrete mixers at the batching plant (which implies how many trucks
can be potentially loaded at the same time),
 distribution type and parameters for truck loading time,
 distribution type and parameters for loaded trucks travel time at the route to each site,
8 Sławomir Biruk

 distribution type and parameters of truck unloading times in each site,

 distribution type and parameters of the time of empty truck return from each site.
Travelling times depend mostly on the distance between the plant and the site, and the
number of crossroads (in urbanized areas). Unloading time depends mostly on the type of
concrete member and methods of delivering the mix from the truck to the formworks (by
means of pumps, crane skips, dumpers etc.).
The second module (Fig. 1) imitates operations of the concrete trucks and concrete
plant. As the shift begins, the trucks (their number is a parameter of the simulation model)
queue at the mixers to be loaded. At this point, destinations of the trucks are decided (as
well as the product they are loaded with – each site may require a different mix); a variety
of strategies may be adopted at this point. For instance, destination of a truck can be chosen
in a random manner, or a truck is sent to a site with a shortest queue. Trucks are loaded
(loading time) and travel to the sites (travel time), where they form a queue waiting to be
unloaded. It was assumed that only one truck can be unloaded at a time. As the unloading
process is completed, the truck returns to the plant and, if the demand of the sites was not
satisfied yet, they queue for another batch of the mix. The procedure is repeated until all
sites receive as much mix as was ordered.

5. Optimizing the allocation of concrete trucks – an example

In the analysed case, one batching plant serves five construction sites. Each site has
a demand for the mix expressed by the number of truckloads (so the assumption was each
batch delivered to each site is of the same volume). These demands are listed in Table 1.
The concreting works at each site start at the same time.

START
Concrete truck are ready

Select
construction site

Queue
to concrete mixer

Truck
loading

Travelling
to construction
site i

Queue
on construction site i

Truck
unloading

Return from
construction
site i

NO Are YES
STOP there any
orders ?

Fig. 1. Diagram of the “simulation engine”

 Dispatching concrete trucks using simulation method 9

The plant is equipped with only one mixer, so that only one truck can be loaded at
a time. It was assumed that the loading time is a random variable of a triangular distribution
defined by the optimistic (ta = 5 min), the pessimistic (tb = 9 min), and the most likely (tm =
6 min) values. It was assumed that the components of the transport cycles are also of
a triangular distribution with the optimistic, pessimistic and most likely values quoted in
Table 1.

Table 1. Input for the example

Demand for the Unloading time, Return time,
Site ID Travel time, minutes
mix, truckloads minutes minutes
1 15 (12,20,15) (8,12,9) (10,17,12)
2 8 (28,45,33) (6,9,7) (26,43,31)
3 12 (22,35,26) (12,18,14) (19,30,24)
4 9 (15,22,17) (9,15,11) (12,20,15)
5 6 (15,25,18) (8,14,10) (14,23,17)

For this simulation, a following strategy was adopted: a loaded truck is directed to the
construction site where the truck queue is the shortest. This strategy minimises the time
since mixing till placing and assures continuity of concreting in particular sites. In the
simulation process, the number of trucks serving the system was change in search for
optimum that assures meeting constraints of concreting process with the least number of
trucks employed. The simulation results are presented in Fig. 2.
If the maximum acceptable time since mixing till placing is 90 minutes (as for stan-
dard mixes at 15°C), the sites must be served by at least 7 trucks. In this case, the batching
plant has to work for 558 minutes to cover the demand.

800

700

600 batching plant

operation time
time [minutes]

500

400
the time since mixing
300 till placing

200
maximum break in
100 the concreting
process
0
5 6 7 8 9 10
number of trucks

Fig. 2. The relationship between the number of trucks and batching plant operation time, maximum break
in the concreting process, and the time since mixing till placing in the analysed case

6. Conclusions
Concrete batching plants are typically equipped with modern automated high-capacity
systems and provide high quality products. The plants often dispose of fleets of concrete
transport trucks. However, the plants rarely use decision support tools to plan transport
routes and schedule production of the mix. The decisions in this respect base on experience
10 Sławomir Biruk

of the staff. The paper investigates into the problems of integrated production scheduling
and vehicle routing. Implementing such systems may increase the number of fulfilled
orders, improve plant and truck utilisation rates and client satisfaction and, as a result, help
to achieve competitive advantage in the highly demanding market of ready mix production.
Further research will be aimed at increasing the number of real-life related constraints
as e.g. changes in transport times during the day related with traffic conditions, allowing for
availability of trucks and pumps of various capacity. This is aimed at increasing reliability
of simulation results.

Acknowledgment
This work was financially supported by Ministry of Science and Higher Education
within the statutory research number S/63/2015

References
1 Asbach L., Dorndorf U., Pesh E. Analysis, modeling and solution of concrete delivery problem.
European Journal of Operational Research 193 (2009) 820–835.
2 Feng Ch. W., Cheng T. M., Wu H. T. Optimizing the schedule of dispatching RMC trucks
through genetic algorithms. Automation in Construction 13 (2004) 327–340.
3 Lin P. C., Wang J., Huang S. H., Wang Y. T. Dispatching ready mixed concrete trucks under
demand postponement and weight limit regulation. Automation in Construction 19 (2010) 798–
807.
4 Lu M., Lam H.-C. Simulation – optimization integrated approach to planning ready mixed
concrete production and delivery: validation and applications. Proceedings of the 2009 Winter
Simulation Conference 2593–2604.
5 Matsatsinis N. F. Towards a decision support system for the ready concrete distribution system:
A case of a Greek company. European Journal of Operational Research 152 (2004) 487–499.
6 Naso D., Surico M., Turchiano B., Kaymak U. Genetic algorithms for supply-chain scheduling:
A case study in the distribution of ready mixed concrete. European Journal of Operational Re-
search 177 (2007) 2069–2099.
7 Park M., Kim W. Y., Lee H. S., Han S. Supply chain model for ready mixed concrete.
Automation in Construction 20 (2011) 44–55.
8 Sobotka A., Biruk S., Jaskowski P. Process approach to production management in
a construction company. Proceeding of the Fourth European Conference on Product and Process
Modelling in the Building and Related Industries, Portož, Slovenia, 19–11 September 2002.
9 Tommelein I. D., Li A. E. Y. Just-in-time concrete delivery: mapping alternatives for vertical
supply chain integration. Proceedings of the Seventh Annual Conference of the International
Group for Lean Construction IGLC-7, California 1999, 97–108.
10 Yan S., Lin H. C., Jiang X. Y. A planning model with a solution algorithm for ready mixed
concrete production and truck dispatching under stochastic travel times. Engineering Optimi-
zation 44 (2012) 427–447.
11 Yan S., Lai W., Chen M. Production scheduling and truck dispatching of ready mixed
concrete. Transportation Research. Part E 44 (2008) 164–179.
12 Yan S., Lin H. Ch., Liu Y.-Ch. Optimal schedule adjustments for supplying ready mixed
concrete following incidents. Automation in Construction 20 (2011) 1041–1050.
13 Yan S., Lai W. An optimal scheduling model for ready mixed concrete supply with overtime
considerations. Automation in Construction 16 (2007) 734–744.
14 Wakefield R., Sears G. Petri Nets for simulation and modeling of construction systems. Journal
of Construction Engineering and Management 123(2) (1997) 105–112.
15 Zayed T.M., Halpin D. Simulation of concrete batch plant production. Journal of Construction
Engineering and Management 127(2) (2001) 132–141.