Human feedback helps control rover navigation with free energy approach
Towards Interaction Regulation from Human Feedback via Free Energy Minimization
Robotics
Summary
Controlling robots or autonomous agents together with humans can be tricky because humans have preferences that may conflict with the robot's goals. The authors propose a new way to include human preferences directly into the robot’s decision-making by using a theory called the free energy principle from brain science. They tested this by letting a person remotely guide a rover using gestures and virtual reality, showing that the system can balance cooperation and competition between the human and the robot’s goals. This method helps the robot adjust its actions based on what the human wants in real time.
What this means in practice
- •For robot operators: Allow remote human operators to guide autonomous rovers with natural gestures that influence the rover’s behavior alongside its own goals.
- •For virtual reality developers: Build VR systems that share sensory data with robots so humans can intuitively provide feedback and influence robot actions in real time.
Authors
Maria Paula Diaz Monfort, Cinzia Tomaselli, Michael Richardson, Giovanni Russo
Abstract
A central challenge across control and learning is the design of mechanisms regulating the interactions between humans and autonomous agents. Inspired by the free energy principle from computational neuroscience, we introduce a control-theoretical framework to integrate human preferences online into an agent policy. We turn the framework into an open control architecture and validate our approach using a human-in-the-loop experimental testbed involving a rover navigating via onboard sensing. The human, remotely located and equipped with virtual reality headsets, shares the same sensory information as the rover. Human preferences are provided to the rover via gestures which introduce both cooperative and competitive interactions between the agent goal and the preferences. The experiments show that interactions are regulated, validating the proposed approach.