Industrial Automation and Robotics Industry Conference

Artificial intelligence is no longer a technology of the future—it is becoming an everyday tool for automation engineers. But how far is industry from factories that not only execute pre-programmed tasks, but also make decisions independently, learn from errors, and adapt to changing conditions?During the presentation, the evolution of AI applications in industry will be outlined—from classical vision systems and predictive maintenance, through intelligent process diagnostics, to the vision of fully autonomous production systems. The role of key technologies will be discussed, including sensor data fusion, real-time image processing, large language models (LLMs) supporting machine operation, and agent-based AI systems capable of autonomous reasoning.The presentation will be illustrated with examples of work carried out at the Silesian University of Technology—from anomaly detection systems in industrial infrastructure using autoencoders and process diagnostics methods, through intelligent gas pipeline monitoring with the use of drones and AI classifiers, to models for predicting energy consumption of manipulators and the implementation of deep learning in autonomous systems operating at the network edge (edge AI). The aim of the presentation is to demonstrate that the path to autonomous production is not a one-time leap, but a process—and that Polish technical universities have a real, measurable contribution to this process.

Artificial intelligence is rapidly transforming industrial automation, but its true value is only realized when AI can operate reliably in the physical world. In this session, Karol Habryka, Director of Application Engineering, EMEA at Teradyne Robotics (Universal Robots), discusses the development of physical artificial intelligence—the next evolution of automation, in which robots not only follow programmed paths, but also perceive their environment, adapt to changing conditions, and act intelligently in real time. Drawing on real-world manufacturing challenges—labor shortages, high-mix production, increasing quality requirements, and constantly changing shop-floor conditions—the speaker explains why traditional automation is reaching its limits and how physical AI enables the emergence of a new generation of adaptive, resilient production systems.

Artificial intelligence is no longer an add-on to automation systems—it is becoming an integral part of them. In this presentation, I will show how advanced digital assistants are redefining the way control systems are designed, commissioned, and maintained, directly addressing the biggest challenges facing today’s industry: time pressure and shortages of skilled personnel. This is not just another chatbot—it is an AI-based agent system capable of generating PLC code, creating HMI visualizations, analyzing live process data, and diagnosing issues, thereby supporting maintenance teams. With this technology, routine tasks are automated, error risks are reduced, and engineers gain time to focus on innovation and solving complex problems. Will AI replace programmers? No. I will demonstrate how this technology enhances your team’s capabilities, enabling rapid machine deployment while keeping full human control and decision-making.

The synergy of advanced 3D vision and artificial intelligence is transforming modern robots into intelligent partners capable of full autonomy, even in unstructured production environments. Contemporary automation emphasizes sustainability, using innovative energy recovery systems to generate tangible operational savings. The latest cybersecurity standards provide the foundation for process protection, ensuring data integrity and production stability in an era of increasing digital threats.

Demographic changes and labor shortages pose significant challenges for the manufacturing industry. In the past, companies could grow by hiring more workers—today, the key is automation using robots. However, many tasks have been difficult or uneconomical to automate due to the required hand-eye coordination, dexterity, and logical thinking. Robotics, combined with artificial intelligence, machine vision, and sensor technology, is now recognized as a key technology enabling effective execution of such tasks. Building on over a century of innovation, Yaskawa presents a new approach in robotics and artificial intelligence, where flexibility meets precision—MOTOMAN NEXT. These new industrial robots are designed to meet the challenges of modern production, from labor shortages to high-mix manufacturing.

Modern industry and maintenance departments face two major challenges: the risk of unexpected failures (including cyberattacks) and an overwhelming amount of complex technical documentation, which takes valuable time to analyze. In this presentation, we will show how advanced data analytics and artificial intelligence models can provide real support to engineers in addressing these problems. The presentation, based on the latest research from the Institute of Automation and Robotics at Warsaw University of Technology, will cover two key areas:

• Early Detection of Hidden Faults and Cyberattacks: We will demonstrate how careful monitoring of controllers and machine behavior can reveal anomalies before they cause production downtime or equipment damage. 
•  Artificial Intelligence in a Sea of Manuals: Can advanced language models (LLMs) quickly identify the cause of a problem and the solution within hundreds of pages of technical documentation? We will present test results of such solutions—highlighting what AI can already do very well and where potential pitfalls and limitations remain.

Robotics doesn’t have to be complicated—provided that the entire process, from planning to maintenance, operates as a single, coherent workflow.

In this presentation, we will demonstrate: 

•  How modern approaches to robot implementation look today 
•  How digital tools and AI shorten production ramp-up time 
•  How to respond more easily to changes 
•  Why simplicity and flexibility are becoming key in industry

The rapid development of artificial intelligence is opening entirely new possibilities for industrial process automation—tasks that until recently seemed achievable only through manual labor. This presentation will showcase the groundbreaking SIMATIC Robot Pick AI solution, which enables robots to pick previously unknown objects of any shape, orientation, or arrangement. Participants will also learn about the extended versions of the system—Pick AI Pro and the currently developing Pack AI—and how leveraging additional visual information (color, depth, segmentation) allows handling more complex grippers and challenging objects. The presentation will demonstrate how artificial intelligence can significantly increase process flexibility, reduce costs, minimize errors, and prepare industrial facilities for the challenges of the future.

Visual quality control has been and will continue to be a key element of virtually every industrial production process. It is worth examining the evolution this technology has undergone in recent years and what it means for the typical user. 

• Two approaches to visual quality control – traditional vs. AI 
• What can be achieved in 3 simple steps, in just a few minutes 
• Examples of applications and use cases

In today’s industry, the question about artificial intelligence is no longer “if,” but “how and where.” Most AI implementations rely on the cloud. But what if the production floor lacks a stable connection? What if process data is too sensitive to leave the facility? What if millisecond delays impact product quality? In this talk, I will show how Polish manufacturing companies can build their own AI infrastructure—from the first pilot to a scalable system—without dependence on external cloud providers. I will discuss real deployment scenarios for AI analytics as well as agent-based systems using LLMs and VLMs, operating locally where the data is generated: directly on the production line.

Artificial intelligence enables a shift from traditional monitoring to advanced predictive methods that genuinely enhance machine reliability. By analyzing data in real time, it becomes possible to detect anomalies and predict potential failures that could halt production. Modern algorithms learn the behavior of equipment, helping to avoid reactive and costly interventions. As a result, companies gain greater operational continuity, lower maintenance costs, and more comprehensive control over their processes.

BLOCK I

The topic refers to the growing role of digital environments in the design, testing, and implementation of industrial systems. Thanks to tools such as digital twins or simulations, engineers can create and optimize production lines without physical access to the factory floor.” “This approach is also changing the profile of automation and robotics engineers—who increasingly need to combine technical expertise with digital and IT competencies, operating in a hybrid environment where the boundary between the physical and virtual worlds is becoming increasingly blurred.

During the presentation, we will show what control system migration looks like in practice in a manufacturing plant or critical infrastructure environment, where there is no room for long downtime, communication errors, or loss of process knowledge. Based on experience from real implementations, we will discuss where risks most often arise, how to prepare an organization for change, and why a well-executed migration does not end with the launch of a new system, but becomes the foundation for further digitalization of production.

The growing need to improve the efficiency of production processes today goes hand in hand with the necessity to reduce downtime, losses, and unplanned failures. Modern manufacturing plants are increasingly seeking solutions that not only improve the performance of machines and installations, but also enable better process control and faster response to emerging problems. During the presentation, we will show what process optimization looks like in practice using industrial automation solutions. Based on concrete case studies from customer implementations, we will discuss the most common challenges encountered in manufacturing plants, as well as the actions and technologies that help increase reliability, operational continuity, and overall production efficiency.

The specific role of SCADA systems as an intermediary between the process layer—controllers, measurement systems, and actuators—and the operator and management layers. SCADA applications originally served for direct supervision of technology but have evolved into systems for data acquisition for production management and enterprise software.

Block II

Presentation of the report by the Sectoral Skills Centre, showing how demand for occupations, qualifications, and skills in the field of robotics is changing. This is not only a diagnosis of skill gaps, but also an attempt to answer the question of how the education system should evolve in order to keep pace with the transformation of industry.

DISCUSSION PANEL

Technology is developing faster than the systems designed to prepare people to work with it.” “A joint discussion among representatives of industry, education, and public institutions on responsibility, cooperation, and the necessary changes. On whether we are able to build a system that not only reacts to change, but begins to stay ahead of it.