Quickie15min
From AI Decisions to Robot Motion: Inside an Industrial Palletizing Stack
An integrated industrial palletizing software stack is presented, connecting AI-based placement decisions, perception, motion planning, robot control, backend orchestration, PLC communication, and web HMI. The architecture shows how reinforcement learning and ROS/MoveIt translate high-level decisions into safe real-robot actions, supported by diagrams, simulations, and demos.
talk.summaryAiDisclaimer
Dimitris KosmasProgressive Robotics
talkDetail.whenAndWhere
Saturday, November 21, 15:35-15:50
Room 1 - Thessaloniki
Designing an integrated software stack for industrial mixed-case palletizing, from AI-based placement decisions to motion planning, robot control, backend orchestration, and a web-based HMI. We will explore the system architecture and how reinforcement learning, perception, ROS/MoveIt, PLC communication, backend services, and 3D visualization work together to turn a high-level decision into a safe action performed by a real industrial robot. The presentation will feature architectural diagrams, simulations, and video demonstrations of individual subsystems.
Dimitris Kosmas
Dimitris Kosmas is a Robotics Engineer at Progressive Robotics, where he works on the development of robotic systems for on-the-fly palletizing solutions. He did a PhD in the Robotics and Mechatronics Group at the University of Twente, where his research focused on developing 3D-printed assistive systems with embedded sensing for orthopedic and gynecological applications. He also holds an MSc in Systems and Control from the same university, specializing in control and compensation methods for systems incorporating 3D-printed sensors.