Salzburg University of Applied Sciences and ABB file joint patent for AI-driven energy optimization in industrial drive systems

Salzburg University of Applied Sciences and ABB's Machine Automation Division (B&R) filed a joint patent application for energy-optimized motion control in industrial automation systems such as robots, machine tools, and automated production lines. The work is anchored in the Josef Ressel Center for Intelligent and Secure Industrial Automation (JRZ ISIA), with research development beginning in 2020 through the EU Interreg project KI-Net and continuing since 2022 within the Center.

The patent applies reinforcement learning (RL)—a method where an AI agent learns directly from a system's real behavior rather than relying solely on mathematical models—to reduce energy losses in highly dynamic motion sequences. A key innovation is a new mathematical formulation that enables faster learning with reduced data requirements, making reinforcement learning methods practical for industrial environments where they were previously considered too slow and data-intensive.

The approach allows control systems to adapt motion profiles autonomously to real operating conditions, with the goal of making motion sequences substantially more energy-efficient while reflecting actual system behavior rather than idealized models.