Breakthrough in Robotics: Integrated Decision-Control Framework Enhances Feasibility of Social Robot Autonomous Navigation

Strategic Insights -

Research from Changchun University of Technology has led to the development of an Integrated Decision-Control Framework for Social Robot Autonomous Navigation (IDC-SRAN), which significantly enhances the feasibility of social robot autonomous navigation. This breakthrough uses reinforcement learning to tackle the challenge of designing pedestrian walking reward and resolves the issue of dynamics mismatch of RL system. The framework enables goal-oriented autonomous navigation through active torque modulation, achieving a task completion rate exceeding 90% and demonstrating peak accelerations approximately 8.3% of baseline methods.

Key Takeaways:

  • The Integrated Decision-Control Framework for Social Robot Autonomous Navigation (IDC-SRAN) is a novel approach to social robot autonomous navigation that accounts for nonlinearity of social robot model and ensures the feasibility of decision-control strategy.
  • IDC-SRAN employs inverse reinforcement learning (IRL) to design pedestrian walking reward and construct a Four-Mecanum-Wheel Robot dynamic model to develop the framework, resolving the issue of dynamics mismatch of RL system.
  • The actions of IDC-SRAN are defined as additional torque, with actual torque and lateral/longitudinal velocities integrated into the state space, and the feasibility of the decision-control strategy is ensured by constraining the range of actions.
  • A driving-force-guided reward is proposed to mitigate the state delay caused by model transient characteristics, which complicates the articulation of nonlinear relationships between states and actions through IRL-based rewards.
  • Experimental results demonstrate that IDC-SRAN achieves a task completion rate exceeding 90% and demonstrates peak accelerations approximately 8.3% of baseline methods.
  • The framework enables goal-oriented autonomous navigation through active torque modulation, significantly enhancing the feasibility of decision-control strategies.
  • The research was financially supported by The Department of Science and Technology of Jilin Province and led by Mingyue Luo from the Changchun University of Technology.

Statistics:

  • 8.3%: The percentage increase in peak accelerations achieved by IDC-SRAN compared to baseline methods.
  • 90%: The task completion rate achieved by IDC-SRAN, exceeding that of baseline methods.
  • 2025: The publication year of the research in PLOS One.
  • 20(6): The volume and issue number of the research in PLOS One.

Sources:

  • PLOS One. "Integrated decision-control for social robot autonomous navigation considering nonlinear dynamics model." PLOS One, vol. 20, no. 6, 2025.
  • Changchun University of Technology. "Integrated Decision-Control for Social Robot Autonomous Navigation (IDC-SRAN)".
  • Mingyue Luo, et al. "Integrated decision-control for social robot autonomous navigation considering nonlinear dynamics model." PLOS One, vol. 20, no. 6, 2025.