Comau AI robotics 2026 is about far more than making robot arms smarter. Across industrial automation, artificial intelligence is now being used to simplify programming, improve robot monitoring, support predictive maintenance, enhance machine vision and make automation more flexible for real production environments.
In this exclusive MachineToolNews.ai interview, Giovanni Di Stefano, Head of Engineering Advanced Robotics at Comau, explains how Comau is applying AI across its robotics and automation portfolio, including MI.RA, in.Grid Robot Monitoring, AI-driven vision systems, condition monitoring and future engineering workflows.
For manufacturers, the key question is no longer whether AI belongs in industrial robotics. It is where AI can deliver measurable value on the shop floor.
Comau AI Robotics 2026: The Interview
MTN: Comau is widely known for industrial robotics, but your recent work shows a strong push into AI. How would you define Comau’s approach to AI in manufacturing today?
Giovanni Di Stefano: Since 2018, Comau has recognized software and AI as core strategic pillars of industrial automation, leading the company to internalize and further develop its digital and engineering know-how. Today, our engineering-driven approach is structured across three main areas. First, we integrate AI directly into standard, scalable solutions such as MI.RA and in.Grid, thus embedding intelligence into robotics, vision systems, and digital platforms. Second, we leverage AI as a key enabler within automation projects, allowing systems to adapt to variability, simplify programming, and support more flexible and dynamic manufacturing environments. Third, we apply AI internally to optimize engineering processes, reduce complexity, and accelerate deployment timelines.
MTN: When manufacturers hear “AI robotics,” many still think of a smarter robot arm. Where are you actually seeing the biggest real-world gains from AI across your portfolio?
Giovanni Di Stefano: The biggest gains, in addition to robot motion, deal with how AI-guidance enhances the overall robustness, adaptability, and effectiveness of automated processes. We are seeing tangible impact in reducing the complexity of programming robotic cells, enabling faster deployment and easier reconfiguration, particularly in high-mix production environments. At the same time, AI-powered vision systems are improving output quality through real-time inspection, defect detection, and adaptive process control.
By collecting and analyzing large volumes of data, solutions such as our in.Grid intelligent platform provide actionable feedback and insights that improve production KPIs, optimize throughput, and support predictive maintenance strategies. This allows operational teams to anticipate failures, minimize downtime, and transition from reactive to predictive operations.
MTN: Comau’s in.Grid Robot Monitoring platform focuses on condition monitoring and performance. In practical shop-floor terms, what does it track, and what decisions does it enable?
Giovanni Di Stefano: in.Grid Robot Monitoring tracks and optimizes process-level performance and overall equipment efficiency within applications such as spot welding, arc welding, and material handling. It also aggregates critical robot KPIs, operational parameters, and performance trends into a unified environment. By leveraging AI-driven advanced analytics and high-frequency data collection, the platform enables faster, more informed decision-making through a lean, cloud-based architecture that connects machines, processes, and operators.
From a practical standpoint, this allows manufacturers to monitor equipment health, identify deviations, optimize cycle times, and improve product quality. At the same time, it supports after-sales teams with remote diagnostics and rapid intervention capabilities, while automatically generating reports for maintenance and production teams.
MTN: At IVECO’s Valladolid plant, this technology is already in use. What has that deployment shown about the real value of AI-driven condition monitoring in production?
Giovanni Di Stefano: The deployment at IVECO’s Valladolid plant has demonstrated that in high-density robotic environments, even marginal efficiency gains translate into significant operational impact.
By consolidating data into intuitive, real-time dashboards, the in.Grid platform has let IVECO quickly identify inefficiencies and has been able to significantly reduce machine downtime and streamline monitoring as a result.
Furthermore, the process-level analysis allows for the optimization of consumables based on actual usage patterns rather than fixed maintenance schedules, improving both cost efficiency and sustainability. Continuous parameter monitoring also enables fine-tuning of production processes, ensuring maximum efficiency and performance.
MTN: Predictive maintenance is often discussed, but rarely quantified. From your perspective, what separates meaningful AI condition monitoring from systems that simply generate more data?
Giovanni Di Stefano: The big difference lies in engineering depth and domain-specific expertise. Comau’s more than 50 years of experience in industrial automation allow us to focus immediately on the variables that truly impact performance and reliability.
Instead of adopting a generic, data-first approach, where large volumes of information are collected without clear prioritization, or using third-party platforms, Comau applies an engineering-driven methodology that targets the specific precursors of critical failure modes. This ensures that the system delivers actionable insights rather than data overload, enabling maintenance teams to intervene proactively and prevent downtime. In other words, we focus on extracting the right data, contextualizing it, and translating it into decisions that improve operational continuity and asset performance.
MTN: With MI.RA and vision-based systems, Comau is enabling robots to handle randomly placed parts without CAD input. How important is that capability for flexible manufacturing environments?
Giovanni Di Stefano: In highly variable environments such as logistics, e-commerce, and advanced manufacturing, traditional programming approaches no longer work due to the unpredictability and diversity of objects and scenarios. For these applications, solutions like MI.RA, which can detect, identify, and autonomously handle randomly placed objects without prior CAD input, are essential.
Without the need for CAD input or fixed positioning, MI.RA allows robots to handle random parts autonomously, providing the essential flexibility required for modern, dynamic manufacturing. This capability enables true flexible and adaptive automation, whereby robots can now operate effectively in unstructured environments where variability is the norm. It also significantly reduces setup time, programming effort, and dependency on specialized skills, making automation more accessible and scalable.
More broadly, this reflects a fundamental shift toward cognitive automation, where robots can perceive their environment, interpret data, and dynamically adjust their actions, which can unlock new levels of productivity, flexibility, and responsiveness.
MTN: How does this kind of AI-driven vision and picking change the economics of automation for manufacturers who may have previously struggled to justify robotics?
Giovanni Di Stefano: There are certain tasks, such as order preparation in logistics or mixed SKU depalletizing, that are impossible without this type of vision, as they require handling a vast number of unknown or variable objects. In other cases, such as depalletizing single SKUs, AI-vision can drastically improve ROI by significantly reducing programming time and simplifying maintenance operations. Furthermore, in traditional manufacturing, vision technology helps reduce costs associated with mechanical fixtures.
Because the vision system compensates for uncertainties, parts no longer need to be perfectly positioned. This means that manufacturers can use simpler racks or bins instead of expensive, precision-machined fixtures, leading to lower initial investment and reduced long-term maintenance costs. It further implies a shift in terms of automation economics because it helps lower entry barriers and total cost of ownership, while enabling applications that were not possible before now.
MTN: Comau has also explored AI in areas such as robot motion optimisation and joint stiffness. How do these advances translate into measurable improvements in accuracy, quality or cycle time?
Giovanni Di Stefano: Merging AI with our control expertise, we have tested the Robot Stiffness Estimator to identify mechanical parameters with a higher level of precision. In fact, by learning the exact correlation between real motor currents and simulated data, we can achieve a more consistent estimation compared to classical methods. This translates into smoother and more controlled robot movements, which significantly reduces mechanical stress. The primary value here is the extended lifespan of the product and improved reliability, as the robot operates under less strain.
Additionally, this AI-driven approach helps streamline tuning and optimization phases, making it faster to reach peak performance while reducing long-term maintenance needs. In practical terms, this ensures more stable process quality, improved repeatability, and optimized cycle times over the lifetime of the system.
MTN: Many manufacturers remain cautious about AI due to cost, complexity and integration challenges. How is Comau making AI-driven automation more accessible and practical to implement?
Giovanni Di Stefano: Our decision to internalize software expertise and develop proprietary platforms allows us to remain hardware-agnostic and highly flexible. This means we can continuously benchmark the best hardware technologies on the market, such as cameras and sensors, and integrate them with minimal effort.
What truly sets Comau apart is our collaborative approach. Our software and Machine Learning engineers work alongside the product engineering and site & commissioning teams, to bring together digital expertise and decades of hands-on robotics experience. This synergy, fueled by a continuous feedback loop from the field, ensures that our advanced AI solutions are also practical, robust, and easy to deploy in real industrial environments. This means that pir customers benefit from solutions that are specifically designed to reduce integration complexity, accelerate deployment, and deliver value quickly.
MTN: Looking ahead, where do you see AI having the biggest impact in industrial robotics over the next three to five years, and what should manufacturers be preparing for now?
Giovanni Di Stefano: We believe AI will deeply transform engineering workflows in the next three to five years by automating repetitive design tasks to enable faster time-to-market and more cost-efficient and sustainable development processes. In parallel, the rise of Large Foundational Models (LFMs), which are essentially broad AI models trained on massive datasets, can be adapted to many downstream tasks and will drastically reduce the time needed to develop custom AI-based solutions. Furthermore, the integration of World Models (AI systems that learn to simulate and predict the physical behavior of their environment) will enable robots to master tasks that are currently impossible, such as the high-speed handling of flexible or deformable objects.
Manufacturers need to prepare for this shift by building a robust and scalable digital infrastructure ready to support these evolving software-driven capabilities. They should also make sure their data management, connectivity, and simulation capabilities can leverage the next generation of AI-enabled automation.
MTN Analysis: AI Robotics Is Moving From Demonstration To Deployment
The strongest message from this interview is that Comau AI robotics 2026 is focused on practical factory outcomes.
Comau is not positioning AI as a standalone technology layer. The company is applying it inside robotics, vision, monitoring and engineering workflows where manufacturers can see a clearer operational benefit.
For metal manufacturers, the most relevant areas are condition monitoring, welding process stability, robotic handling, machine vision and flexible automation. These are the points where AI can make a measurable difference to uptime, quality, maintenance planning and labour efficiency.
The IVECO Valladolid deployment is especially important because it shows AI robot monitoring moving into a live production environment. That kind of reference point matters for manufacturers that want evidence before committing to broader AI adoption.
The next stage will be integration. AI robotics will become more valuable when vision, monitoring, simulation, maintenance and process control work together inside connected production cells.
Key Takeaways
Comau AI robotics 2026 is focused on practical factory outcomes, including faster deployment, predictive maintenance, vision-guided picking and improved production visibility.
Comau’s in.Grid Robot Monitoring platform is designed to help manufacturers track robot performance, identify deviations and support maintenance decisions.
MI.RA vision systems allow robots to handle more variability, including randomly placed parts and applications without prior CAD input.
AI-driven vision can improve the economics of automation by reducing programming time, lowering fixture costs and making robotics suitable for more applications.
Manufacturers should prepare for AI robotics by improving data infrastructure, connectivity and simulation readiness.
FAQ
What is Comau AI robotics 2026?
Comau AI robotics 2026 refers to Comau’s use of artificial intelligence across industrial robotics, vision systems, condition monitoring, robot monitoring, predictive maintenance and engineering workflows.
What does Comau’s in.Grid Robot Monitoring platform do?
Comau’s in.Grid Robot Monitoring platform tracks robot performance, process data, operational parameters and equipment health. It helps manufacturers identify deviations, optimise cycle times and support predictive maintenance decisions.
How is Comau using AI in machine vision?
Comau uses AI in MI.RA vision systems to help robots detect, identify and handle objects in more variable environments. This includes randomly placed parts and applications where traditional CAD-based programming may be too slow or restrictive.
Why does AI vision improve automation ROI?
AI vision can reduce programming time, simplify maintenance and lower the need for expensive fixtures. It allows robots to deal with uncertainty, which makes automation more flexible and easier to justify.
What should manufacturers do before adopting AI robotics?
Manufacturers should improve their data management, connectivity, simulation tools and digital infrastructure. These foundations will make it easier to adopt AI-enabled automation and gain measurable value from it.
Further Reading On MachineToolNews.ai
What Is Physical AI in Robotics and Automation?
Agile Robots Physical AI Robotics: 5 Insights from Sven Parusel
What Is Industrial AI and How Is It Used in Factories?
AI in Machine Tools 2026: 10 Breakthrough Technologies Changing Manufacturing
