Toward Safer Navigation of Heterogeneous Mobile Robots in Distributed Scheme: A Novel Time-to-Collision-Based Method.

Area of Science:

• Robotics
• Control Systems
• Artificial Intelligence

Background:

• Safe navigation for heterogeneous multiple mobile robots (HMRs) requires considering robot dynamics (mass, inertia).
• Existing methods often rely on kinematics or point-mass models, leading to conservative or failed navigation, particularly with robots of differing masses.
• This limitation hinders efficient operation in complex environments like warehouses.

Purpose of the Study:

• To develop a novel, dynamics-aware navigation methodology for HMRs.
• To design a distributed control scheme that explicitly incorporates robot dynamics to prevent collisions.
• To enhance the safety and efficiency of HMR navigation, especially in scenarios with significant mass/inertia differences.

Main Methods:

• Proposed a new predictive collision term based on time to collision (TTC) to quantify imminent collisions.
• Developed a novel nonlinear controller that integrates the TTC term for collision-free motion guarantees.
• Validated the approach through simulations and experiments, comparing it against kinematics-only controllers.

Main Results:

• The proposed TTC-based controller demonstrated less conservative safe motion planning compared to kinematics-only methods.
• Kinematics-based approaches resulted in collisions in confined spaces, whereas the TTC strategy ensured safe navigation.
• Experimental validation confirmed collision-free navigation for HMRs using the developed method.

Conclusions:

• The developed TTC-based control strategy provides more reliable motion control for HMRs, particularly those with significant differences in mass or inertia.
• This approach is beneficial for applications like warehouse automation and autonomous driving, improving collision avoidance.
• The methodology offers a significant advancement in ensuring safe and efficient multi-robot systems operation.