AUSIL SYSTEMS

[𝗘𝗡𝗚] We are delighted to share some great news: Ausil Systems will be present at the next edition of the K – The World's No.1 Trade Fair for Plastics and Rubber in Düsseldorf, the largest international gathering for the plastics and rubber industry. We will be attending alongside our partners, with whom we continue to build innovative solutions for the thermoforming sector. It will be a unique opportunity to showcase our technologies, discover new trends and, above all, connect with customers and professionals from around the world. 👉 The event will take place in October 2025, and we are already preparing everything to ensure that our participation reflects the spirit that drives us: innovation, trust, and collaboration. See you in Düsseldorf! [𝗘𝗦𝗣] Nos alegra compartir una gran noticia: Ausil Systems estará presente en la próxima edición de la feria K – The World's No.1 Trade Fair for Plastics and Rubber en Düsseldorf, el mayor encuentro internacional para la industria del plástico y el caucho. Acudiremos acompañando a nuestros partners, con quienes seguimos construyendo soluciones innovadoras para el sector del termoformado. Será una ocasión única para mostrar nuestras tecnologías, descubrir nuevas tendencias y, sobre todo, conectar con clientes y profesionales de todo el mundo. 👉 La cita será en octubre de 2025 y ya estamos preparando todo para que nuestra participación refleje el espíritu que nos mueve: innovación, confianza y colaboración. ¡Nos vemos en Düsseldorf!

When “normal” is the most valuable: 𝗔𝗻𝗼𝗺𝗮𝗹𝘆 𝗗𝗲𝘁𝗲𝗰𝘁𝗶𝗼𝗻 𝘄𝗶𝘁𝗵 𝗔𝗜 𝗶𝗻 𝗧𝗵𝗲𝗿𝗺𝗼𝗳𝗼𝗿𝗺𝗶𝗻𝗴 In the thermoforming industry, defects can appear at any moment… and in any form. Some are well-known (cracks, bubbles), others show up unexpectedly: a grease spot, a hair, a water stain. How can we detect them all without training the AI with thousands of different defect examples? That’s where Deep Learning-based Anomaly Detection models come into play. Instead of training a neural network to recognize every possible defect, we train the system to learn what a compliant product looks like. Anything that deviates from that “normality” is flagged as suspicious. In our case, we use architectures such as Autoencoders, GANs, or state-of-the-art techniques like 𝗥𝗲𝘃𝗲𝗿𝘀𝗲 𝗗𝗶𝘀𝘁𝗶𝗹𝗹𝗮𝘁𝗶𝗼𝗻 or 𝗙𝗮𝘀𝘁𝗙𝗹𝗼𝘄, and we’ve applied this technology directly on thermoforming lines. High-resolution cameras capture every thermoformed part, and the neural network returns an “anomaly score” that enables real-time decisions: reject the unit, alert the operator, or trigger a secondary inspection. Key advantages: ✅ Detects rare or unknown defects never seen during training. ✅ Reduces the need for labeling large volumes of defective images. ✅ Simplifies tasks and speeds up commissioning of the vision system. ✅ Provides comprehensive monitoring of each tray’s surface quality. And yes, there are challenges: 🔘 Defining what’s truly “normal” (minor acceptable variations can be mistaken for defects). 🔘 It doesn’t classify the type of defect—just signals that something is off. 🔘 Requires fine-tuning the sensitivity threshold to avoid false positives. Our latest results using 𝗥𝗲𝘃𝗲𝗿𝘀𝗲 𝗗𝗶𝘀𝘁𝗶𝗹𝗹𝗮𝘁𝗶𝗼𝗻 have been promising. But we keep working to make the system better: more good images, less “line noise,” and hybrid models tailored to defect types. At Ausil Systems, we believe AI-based anomaly detection is a powerful tool to ensure quality in highly variable environments. If you’re also exploring new ways to guarantee production excellence, we’d love to exchange insights! #IndustrialAI #Thermoforming #QualityControl #MachineVision #AnomalyDetection #SmartManufacturing #RPCONTROL #AusilSystems

Cuando lo “normal” es lo más valioso: 𝗗𝗲𝘁𝗲𝗰𝗰𝗶ó𝗻 𝗱𝗲 𝗔𝗻𝗼𝗺𝗮𝗹í𝗮𝘀 𝗰𝗼𝗻 𝗜𝗔 𝗲𝗻 𝗧𝗲𝗿𝗺𝗼𝗳𝗼𝗿𝗺𝗮𝗱𝗼 En la industria del termoformado, los defectos pueden surgir en cualquier momento… y en cualquier forma. Algunos son conocidos (grietas, burbujas), otros aparecen por primera vez sin previo aviso: una mota de grasa, un cabello, una mancha de agua. ¿Cómo detectarlos todos sin entrenar la IA con miles de ejemplos distintos? Ahí es donde entran los modelos de Detección de Anomalías basados en Deep Learning. En lugar de entrenar una red neuronal para detectar cada tipo de defecto, entrenamos al sistema para aprender qué es un 𝗽𝗿𝗼𝗱𝘂𝗰𝘁𝗼 𝗰𝗼𝗻𝗳𝗼𝗿𝗺𝗲. Todo lo que se desvíe de esa "normalidad" se considera sospechoso. En nuestro caso, usamos arquitecturas como Autoencoders, GANs o técnicas de última generación como 𝗥𝗲𝘃𝗲𝗿𝘀𝗲 𝗗𝗶𝘀𝘁𝗶𝗹𝗹𝗮𝘁𝗶𝗼𝗻 o 𝗙𝗮𝘀𝘁𝗙𝗹𝗼𝘄, y hemos aplicado esta tecnología directamente en líneas de termoformado. Cámaras de alta resolución capturan cada termoformado, y la red neuronal devuelve un “anomaly score” que permite tomar decisiones en tiempo real: rechazar la unidad, alertar al operario o iniciar una inspección secundaria. 𝗩𝗲𝗻𝘁𝗮𝗷𝗮𝘀 𝗰𝗹𝗮𝘃𝗲: ✅ Detecta 𝗱𝗲𝗳𝗲𝗰𝘁𝗼𝘀 𝗿𝗮𝗿𝗼𝘀 𝗼 𝗱𝗲𝘀𝗰𝗼𝗻𝗼𝗰𝗶𝗱𝗼𝘀 nunca vistos durante el entrenamiento. ✅𝗥𝗲𝗱𝘂𝗰𝗲 𝗹𝗮 𝗻𝗲𝗰𝗲𝘀𝗶𝗱𝗮𝗱 𝗱𝗲 𝗲𝘁𝗶𝗾𝘂𝗲𝘁𝗮𝗿 grandes volúmenes de imágenes con defectos. ✅𝗦𝗶𝗺𝗽𝗹𝗶𝗳𝗶𝗰𝗮 𝗹𝗮𝘀 𝘁𝗮𝗿𝗲𝗮𝘀 𝘆 𝗿𝗲𝗱𝘂𝗰𝗲 𝗹𝗼𝘀 𝘁𝗶𝗲𝗺𝗽𝗼𝘀 de puesta en marcha de la máquina de visión. ✅Supervisa de forma integral la 𝗰𝗮𝗹𝗶𝗱𝗮𝗱 𝘀𝘂𝗽𝗲𝗿𝗳𝗶𝗰𝗶𝗮𝗹 de cada bandeja. Y también hay retos: 🔘 Definir bien qué es “normal” (variaciones menores aceptables pueden confundirse con defectos). 🔘No clasifica el tipo de defecto, solo indica que algo no encaja. 🔘Requiere un fino ajuste del umbral de sensibilidad para evitar falsos positivos. Nuestros resultados más recientes, utilizando 𝗥𝗲𝘃𝗲𝗿𝘀𝗲 𝗗𝗶𝘀𝘁𝗶𝗹𝗹𝗮𝘁𝗶𝗼𝗻, han sido prometedores. Pero seguimos trabajando para mejorar el sistema: más imágenes buenas, menos “ruido de línea”, y modelos híbridos por tipo de defecto. En Ausil Systems creemos que la detección de anomalías con IA es una herramienta poderosa para garantizar calidad en entornos con alta variabilidad. Y si tú también estás explorando nuevas formas de asegurar la excelencia en producción, ¡nos encantará compartir experiencias! #IAIndustrial #Termoformado #ControlDeCalidad #VisiónArtificial #AnomalyDetection #ManufacturaInteligente #RPCONTROL #AusilSystems

🚀 On May 21st, Antonio Pérez, CTO of AUSIL SYSTEMS, took the stage at the SPE Thermoforming Division Conference in Atlanta, proudly representing IFS - Integrated Forming Solutions. 🎙 His presentation, 𝗔𝗜 & 𝗗𝗶𝗴𝗶𝘁𝗮𝗹𝗶𝘇𝗮𝘁𝗶𝗼𝗻 𝗶𝗻 𝗧𝗵𝗲𝗿𝗺𝗼𝗳𝗼𝗿𝗺𝗶𝗻𝗴: 𝗧𝗵𝗲 𝗡𝗲𝘅𝘁 𝗙𝗿𝗼𝗻𝘁𝗶𝗲𝗿, explored how Digital Twins, machine learning, and real-time data integration are transforming traditional thermoforming into smart, data-driven processes. These technologies are enabling manufacturers to achieve greater process control, predictive quality assurance, and operational efficiency—paving the way for Thermoforming 4.0. 📍Location: Cobb Galleria Centre, Atlanta, Georgia 👥 Attendance: Around 150 professionals, including engineers, executives, researchers, and OEMs from across the thermoforming industry. Antonio also shared practical insights from recent projects within the 𝗜𝗙𝗦 𝗚𝗿𝗼𝘂𝗽, including real-world use cases where our AI-powered inspection systems and digital thread strategies have helped clients cut scrap by 50% and reduce defective products delivered to customers by 99.9%. Beyond the technical discussions, the event was a fantastic opportunity to: 🔗 Connect with key players in the thermoforming ecosystem 🧠 Exchange ideas on the role of AI, sustainability, and automation in the future of manufacturing 🌍 Showcase how our technology is driving the development of smarter, more efficient, and more sustainable production environments Huge thanks to the SPE Thermoforming Division for hosting such a forward-looking event, and to everyone who joined us in shaping the future of the industry. Let’s keep innovating, collaborating, and pushing the boundaries of thermoforming together! #Thermoforming #AI #DigitalTwins #Industry40 #SmartManufacturing #IFSGroup #AusilSystems #Innovation #SPE2025 #Automation #Sustainability

📹 𝗖𝘂𝗮𝗻𝗱𝗼 𝗹𝗮𝘀 𝗰á𝗺𝗮𝗿𝗮𝘀 𝗱𝗲𝗷𝗮𝗻 𝗱𝗲 𝗺𝗶𝗿𝗮𝗿 𝘆 𝗰𝗼𝗺𝗶𝗲𝗻𝘇𝗮𝗻 𝗮 𝗲𝗻𝘁𝗲𝗻𝗱𝗲𝗿   Durante años, las cámaras de seguridad han sido testigos silenciosos. Han grabado, archivado y esperado a que algo pase… o mejor dicho, a que alguien revise. Son ojos sin cerebro, acumulando miles de horas de vídeo que, en la práctica, rara vez se revisan. Hasta ahora.   Hoy, las cámaras ya no solo ven: 𝗲𝗻𝘁𝗶𝗲𝗻𝗱𝗲𝗻. Y eso cambia radicalmente su papel en las empresas.   Ya no hablamos solo de detectar intrusiones fuera de horario. Ahora pueden identificar comportamientos anómalos definidos previamente. Pueden 𝘀𝗲𝗴𝘂𝗶𝗿 𝗮 𝘂𝗻𝗮 𝗽𝗲𝗿𝘀𝗼𝗻𝗮 𝗮 𝘁𝗿𝗮𝘃é𝘀 𝗱𝗲 𝗱𝗶𝘀𝘁𝗶𝗻𝘁𝗮𝘀 𝘇𝗼𝗻𝗮𝘀 𝗱𝗲 𝘂𝗻𝗮 𝗽𝗹𝗮𝗻𝘁𝗮, 𝗮𝗻𝗮𝗹𝗶𝘇𝗮𝗿 𝘀𝗶 𝗮𝗹𝗴𝘂𝗶𝗲𝗻 𝗲𝘀𝘁á 𝘂𝘀𝗮𝗻𝗱𝗼 𝘀𝘂 𝗲𝗾𝘂𝗶𝗽𝗼 𝗱𝗲 𝗽𝗿𝗼𝘁𝗲𝗰𝗰𝗶ó𝗻, 𝘀𝗶 𝘀𝗲 𝘂𝘀𝗮𝗻 𝗱𝗶𝘀𝗽𝗼𝘀𝗶𝘁𝗶𝘃𝗼𝘀 𝗻𝗼 𝗮𝘂𝘁𝗼𝗿𝗶𝘇𝗮𝗱𝗼𝘀, como móviles o cámaras, 𝗱𝗲𝘁𝗲𝗰𝘁𝗮𝗿 𝘂𝗻𝗮 𝗰𝗮í𝗱𝗮, o incluso 𝗮𝗹𝗲𝗿𝘁𝗮𝗿 𝘀𝗶 𝘂𝗻 𝗼𝗽𝗲𝗿𝗮𝗿𝗶𝗼 𝗵𝗮 𝗮𝗰𝗰𝗲𝗱𝗶𝗱𝗼 𝗮 𝘂𝗻𝗮 𝘇𝗼𝗻𝗮 𝗿𝗲𝘀𝘁𝗿𝗶𝗻𝗴𝗶𝗱𝗮. ⚠️   Lo más interesante no es la tecnología en sí, sino lo que representa: 𝘂𝗻𝗮 𝗻𝘂𝗲𝘃𝗮 𝗺𝗮𝗻𝗲𝗿𝗮 𝗱𝗲 𝗲𝗻𝘁𝗲𝗻𝗱𝗲𝗿 𝗹𝗮 𝘀𝗲𝗴𝘂𝗿𝗶𝗱𝗮𝗱, 𝗹𝗮 𝗲𝗳𝗶𝗰𝗶𝗲𝗻𝗰𝗶𝗮 𝘆 𝗹𝗮 𝘀𝘂𝗽𝗲𝗿𝘃𝗶𝘀𝗶ó𝗻 𝗼𝗽𝗲𝗿𝗮𝘁𝗶𝘃𝗮.   En lugar de tener a un vigilante mirando veinte pantallas, tienes una IA que filtra lo importante y lo urgente. Que no se cansa, que no se distrae, que no necesita dormir.   Esto 𝗻𝗼 𝘀𝘂𝘀𝘁𝗶𝘁𝘂𝘆𝗲 𝗽𝗲𝗿𝘀𝗼𝗻𝗮𝘀, 𝗺𝗲𝗷𝗼𝗿𝗮 𝗽𝗿𝗼𝗰𝗲𝘀𝗼𝘀. El verdadero valor está en cómo se integran estos sistemas con los flujos de trabajo del día a día: alertas que llegan al móvil de un supervisor, informes de presencia en zonas críticas, evidencias en tiempo real para actuar antes de que el problema escale.   Pero lo que es evidente es que la videovigilancia ya no es lo que era. Se ha convertido en una herramienta inteligente, y como tal, merece una conversación seria sobre su rol en las empresas del presente.   No se trata de ver más. Se trata de ver mejor. Y sobre todo, de entender a tiempo. #visionartifical #seguridad #inteligenciaartificial

𝗥𝗲𝗮𝗹-𝗧𝗶𝗺𝗲 𝗜𝗻𝘀𝗽𝗲𝗰𝘁𝗶𝗼𝗻 𝗳𝗼𝗿 𝗣𝗿𝗲𝗰𝗶𝘀𝗲 𝗤𝘂𝗮𝗹𝗶𝘁𝘆 𝗖𝗼𝗻𝘁𝗿𝗼𝗹 In thermoforming, even the smallest flaw can disrupt production, lead to material waste, and affect customer satisfaction. At AUSIL SYSTEMS, we developed 𝗥𝗣𝗖𝗢𝗡𝗧𝗥𝗢𝗟 to prevent that. This AI-powered vision system inspects thermoformed packaging in real time, identifying defects and helping operators take immediate action. We’re proud to introduce a major upgrade: 𝘁𝗵𝗲 𝗥𝗣𝗖𝗢𝗡𝗧𝗥𝗢𝗟 𝗟𝗶𝘃𝗲 𝗗𝗮𝘀𝗵𝗯𝗼𝗮𝗿𝗱. This real-time interface gives manufacturers complete visibility over their line—showing performance stats, defect types, and system recommendations on a single screen. Key insights provided include:  🔍 𝗧𝗼𝘁𝗮𝗹 𝗶𝗻𝘀𝗽𝗲𝗰𝘁𝗶𝗼𝗻𝘀 in real time 📊 𝗕𝗿𝗲𝗮𝗸𝗱𝗼𝘄𝗻 𝗼𝗳 𝗱𝗲𝗳𝗲𝗰𝘁 𝘁𝘆𝗽𝗲𝘀 (cutting, contamination, etc.) 📈 𝗗𝗲𝗳𝗲𝗰𝘁 𝘁𝗿𝗲𝗻𝗱𝘀 for process improvement ⚠️ 𝗟𝗶𝘃𝗲 𝗮𝗹𝗲𝗿𝘁𝘀 with actionable suggestions What makes this dashboard different is its instant feedback—it turns data into decisions. By doing so, 𝗥𝗣𝗖𝗢𝗡𝗧𝗥𝗢𝗟 becomes more than a quality gatekeeper; it’s a tool to enhance productivity and minimize losses across production runs. How are you using real-time data to improve quality in your processes? #computervision #industry40 #thermoforming #qualitycontrol #smartmanufacturing

📉 𝗔𝘃𝗼𝗶𝗱𝗶𝗻𝗴 𝗣𝗮𝗰𝗸𝗮𝗴𝗶𝗻𝗴 𝗘𝗿𝗿𝗼𝗿𝘀: 𝗧𝗵𝗲 𝗞𝗲𝘆 𝘁𝗼 𝗘𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝘁 𝗣𝗿𝗼𝗱𝘂𝗰𝘁𝗶𝗼𝗻 In thermoforming lines, every detail matters. A small defect in a tray might seem insignificant, but the consequences can be huge. A faulty seal, deformation, or contamination that goes undetected can lead to customer complaints, material waste, and unexpected production downtime. ⚠️ The impact goes beyond the cost of a single defective batch. Companies can face 𝗳𝗶𝗻𝗮𝗻𝗰𝗶𝗮𝗹 𝗹𝗼𝘀𝘀𝗲𝘀 due to returns, regulatory non-compliance, or reputational damage. In a competitive market, maintaining flawless quality control isn’t optional—it’s essential. This is where 𝗮𝗿𝘁𝗶𝗳𝗶𝗰𝗶𝗮𝗹 𝗶𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝗰𝗲 is making a real difference. With automated inspection systems, every tray can be analyzed in 𝗿𝗲𝗮𝗹-𝘁𝗶𝗺𝗲, with an accuracy far beyond manual visual checks. The result? 𝗠𝗼𝗿𝗲 𝗲𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝗰𝘆, 𝗹𝗲𝘀𝘀 𝘄𝗮𝘀𝘁𝗲, 𝗮𝗻𝗱 𝗮 𝗺𝘂𝗰𝗵 𝗵𝗶𝗴𝗵𝗲𝗿 𝗾𝘂𝗮𝗹𝗶𝘁𝘆 𝗴𝘂𝗮𝗿𝗮𝗻𝘁𝗲𝗲. ✅ Factories that have adopted this technology are already seeing the benefits: 𝗳𝗲𝘄𝗲𝗿 𝗱𝗲𝗳𝗲𝗰𝘁𝘀, 𝗶𝗻𝗰𝗿𝗲𝗮𝘀𝗲𝗱 𝗽𝗿𝗼𝗱𝘂𝗰𝘁𝗶𝘃𝗶𝘁𝘆, 𝗮𝗻𝗱 𝗳𝘂𝗹𝗹 𝗰𝗼𝗻𝘁𝗿𝗼𝗹 𝗼𝘃𝗲𝗿 𝗲𝘃𝗲𝗿𝘆 𝘂𝗻𝗶𝘁 𝗽𝗿𝗼𝗱𝘂𝗰𝗲𝗱. With advanced 𝗮𝗻𝗮𝗹𝘆𝘀𝗶𝘀 𝗮𝗻𝗱 𝘁𝗿𝗮𝗰𝗲𝗮𝗯𝗶𝗹𝗶𝘁𝘆 𝘁𝗼𝗼𝗹𝘀, AI doesn’t just detect errors—it helps prevent them before they happen. The real question isn’t whether your production line needs inspection, but whether it’s ready to take the next step toward 𝗶𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝘁 𝗾𝘂𝗮𝗹𝗶𝘁𝘆 𝗰𝗼𝗻𝘁𝗿𝗼𝗹. 📩 𝗪𝗮𝗻𝘁 𝘁𝗼 𝘀𝗲𝗲 𝗵𝗼𝘄 𝗔𝗜-𝗽𝗼𝘄𝗲𝗿𝗲𝗱 𝗶𝗻𝘀𝗽𝗲𝗰𝘁𝗶𝗼𝗻 𝗰𝗮𝗻 𝘁𝗿𝗮𝗻𝘀𝗳𝗼𝗿𝗺 𝘆𝗼𝘂𝗿 𝗽𝗿𝗼𝗱𝘂𝗰𝘁𝗶𝗼𝗻 𝗹𝗶𝗻𝗲? 𝗖𝗼𝗻𝘁𝗮𝗰𝘁 𝘂𝘀 𝘁𝗼𝗱𝗮𝘆 𝗮𝗻𝗱 𝗹𝗲𝘁’𝘀 𝘁𝗮𝗸𝗲 𝗾𝘂𝗮𝗹𝗶𝘁𝘆 𝘁𝗼 𝘁𝗵𝗲 𝗻𝗲𝘅𝘁 𝗹𝗲𝘃𝗲𝗹! #Thermoforming #AI #QualityControl #Industry40 #Automation #SmartManufacturing

🔥 𝗧𝗵𝗲𝗿𝗺𝗼𝗳𝗼𝗿𝗺𝗶𝗻𝗴 𝗶𝘀 𝗯𝗼𝗼𝗺𝗶𝗻𝗴, 𝗮𝗻𝗱 𝘄𝗶𝘁𝗵 𝗶𝘁 𝗰𝗼𝗺𝗲𝘀 𝘁𝗵𝗲 𝗻𝗲𝗲𝗱 𝘁𝗼 𝗲𝘃𝗼𝗹𝘃𝗲 The global thermoformed packaging market is on an upward trajectory, projected to reach $87.73 billion by 2033, with a steady annual growth rate of 5.85%. As industries such as food and beverage, pharmaceuticals, and electronics continue to demand more versatile, efficient, and sustainable packaging solutions, the sector faces challenges that cannot be ignored. The increasing pressure to ensure high-quality standards, comply with stricter environmental regulations, and optimize operational costs without compromising efficiency has made it clear that traditional inspection and quality control processes are no longer enough. In this context, artificial intelligence is not just an innovation—it is an essential tool for the future of thermoforming. In-line inspection powered by AI and computer vision enables real-time defect detection, eliminating human errors and ensuring that only products meeting the highest standards reach the market. Beyond improving quality, this technology reduces material waste, optimizes production speed, and guarantees regulatory compliance without relying on costly, repetitive manual inspections. But the benefits go even further. As companies strive to adopt more sustainable practices and reduce their environmental impact, AI plays a key role by minimizing defects, maximizing resource efficiency, and contributing to a more responsible and streamlined production process. The growth of the thermoformed packaging market presents a great opportunity, but also a challenge: companies that fail to embrace advanced technologies risk falling behind in an increasingly demanding and competitive industry. The question is not whether AI will be part of the future of thermoforming, but rather who will take the lead in this transformation. 💡 𝗜𝘀 𝘆𝗼𝘂𝗿 𝗰𝗼𝗺𝗽𝗮𝗻𝘆 𝗿𝗲𝗮𝗱𝘆 𝗳𝗼𝗿 𝘁𝗵𝗶𝘀 𝗲𝘃𝗼𝗹𝘂𝘁𝗶𝗼𝗻? If you want to explore how artificial intelligence can optimize your thermoforming line, let’s talk. 🚀 #Thermoforming #ArtificialIntelligence #Industry40