Sign in Subscribe
Machine learning

Novel Algorithm Improves Motion Learning in Robots

Strategic Insights

— 1 min read

Researchers from Technology and Innovation Centre have proposed a new algorithm called Constrained Expectation Maximization (CEM) to address a limitation in teaching motion skills to robots through demonstrations. The CEM algorithm enforces time-sensitive constraints (TSC) during the learning process, enabling improved and more efficient reproduction of demonstration data. This breakthrough has significant implications for industrial applications of robotics, where precise execution of motion is crucial.

Key Takeaways:

  • The CEM algorithm is a novel approach to task-parameterized motion learning, which allows robots to learn complex motion skills through demonstrations.
  • The algorithm enforces time-sensitive constraints (TSC) during the learning process, ensuring that the robot accurately reproduces the demonstration data.
  • The CEM algorithm has been tested and validated on handwritten data and real robot applications using the KUKA LBR iiwa, demonstrating its effectiveness in improving motion learning.
  • The research was funded by KUKA Deutschland GmbH, the State of Bavaria through the OPERA, and the Fundacao para a Ciencia e a Tecnologia (FCT).
  • The study highlights the importance of precise execution of motion in industrial applications, where the CEM algorithm's ability to enforce time-sensitive constraints is crucial.
  • The team, led by Julian Richter, Kuka, Technology and Innovation Centre, includes additional authors Christian Scheurer, Niels Dehio, Jochen J. Steil, and Joao Oliveira.

Statistics:

  • The CEM algorithm was tested on handwritten data and real robot applications using the KUKA LBR iiwa.
  • The study demonstrated that the CEM algorithm achieves improved and more efficient reproduction of demonstration data compared to state-of-the-art methods.
  • The research was published in Ieee Robotics and Automation Letters, a peer-reviewed journal.
  • The paper "Task-parameterized Motion Learning With Time-sensitive Constraints" is available online through Ieeexplore.
  • The research was funded by the State of Bavaria through the OPERA, which allocated funds for the project.

Sources:

  • Task-parameterized Motion Learning With Time-sensitive Constraints. Ieee Robotics and Automation Letters, 2025;10(11):11427-11434.
  • VerticalNews. "Researchers detail new data in Robotics - Robotics and Automation." 2025-11-03.
  • Ieee-inst Electrical Electronics Engineers Inc. 445 Hoes Lane, Piscataway, NJ 08855-4141, USA.
  • Technology and Innovation Centre, Kuka. D-86165 Augsburg, Germany.

Read more