Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Edge Accelerated Robot Navigation With Collaborative Motion Planning

Guoliang Li, Ruihua Han, Shuai Wang, Fei Gao, Yonina C. Eldar, Chengzhong Xu

TL;DR

Experiments show that EARN achieves significantly smaller navigation time and higher success rates than state-of-the-art navigation approaches and is validated in indoor simulation, outdoor simulation, and real-world environments.

Abstract

Low-cost distributed robots suffer from limited onboard computing power, resulting in excessive computation time when navigating in cluttered environments. This paper presents Edge Accelerated Robot Navigation (EARN), to achieve real-time collision avoidance by adopting collaborative motion planning (CMP). As such, each robot can dynamically switch between a conservative motion planner executed locally to guarantee safety (e.g., path-following) and an aggressive motion planner executed non-locally to guarantee efficiency (e.g., overtaking). In contrast to existing motion planning approaches that ignore the interdependency between low-level motion planning and high-level resource allocation, EARN adopts model predictive switching (MPS) that maximizes the expected switching gain with respect to robot states and actions under computation and communication resource constraints. The MPS problem is solved by a tightly-coupled decision making and motion planning framework based on bilevel mixed-integer nonlinear programming and penalty dual decomposition. We validate the performance of EARN in indoor simulation, outdoor simulation, and real-world environments. Experiments show that EARN achieves significantly smaller navigation time and higher success rates than state-of-the-art navigation approaches.

Edge Accelerated Robot Navigation With Collaborative Motion Planning

TL;DR

Experiments show that EARN achieves significantly smaller navigation time and higher success rates than state-of-the-art navigation approaches and is validated in indoor simulation, outdoor simulation, and real-world environments.

Abstract

Low-cost distributed robots suffer from limited onboard computing power, resulting in excessive computation time when navigating in cluttered environments. This paper presents Edge Accelerated Robot Navigation (EARN), to achieve real-time collision avoidance by adopting collaborative motion planning (CMP). As such, each robot can dynamically switch between a conservative motion planner executed locally to guarantee safety (e.g., path-following) and an aggressive motion planner executed non-locally to guarantee efficiency (e.g., overtaking). In contrast to existing motion planning approaches that ignore the interdependency between low-level motion planning and high-level resource allocation, EARN adopts model predictive switching (MPS) that maximizes the expected switching gain with respect to robot states and actions under computation and communication resource constraints. The MPS problem is solved by a tightly-coupled decision making and motion planning framework based on bilevel mixed-integer nonlinear programming and penalty dual decomposition. We validate the performance of EARN in indoor simulation, outdoor simulation, and real-world environments. Experiments show that EARN achieves significantly smaller navigation time and higher success rates than state-of-the-art navigation approaches.
Paper Structure (17 sections, 24 equations, 13 figures, 3 tables, 2 algorithms)

This paper contains 17 sections, 24 equations, 13 figures, 3 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Related Work
  3. Tightly Coupled Decision-Making and Motion Planning
  4. Collaborative Motion Planning
  5. Model Predictive Switching
  6. Case Study
  7. Algorithm Designs for Decision-Making and Motion Planning
  8. Decision-Making
  9. Motion Planning
  10. Experiments
  11. Implementation
  12. Benchmarks
  13. Single-Robot Simulation
  14. Multi-Robot Simulation
  15. Intel CARLA Challenge
  16. ...and 2 more sections

Figures (13)

  • Figure 1: Low-cost robots execute shape distance collision avoidance assisted by a proximal edge server based on EARN.
  • Figure 2: Architecture of EARN.
  • Figure 3: Computation time (ms) of PDD planner versus the number of prediction horizons $H$ and obstacles $|\mathcal{M}(k)|$.
  • Figure 4: Crossroad scenario in CARLA Town04 and planner switching from PF to PDD performed by EARN.
  • Figure 5: Trajectories and control parameters of different schemes in crossroad scenario. Trajectories generated by the local and edge motion planners are marked in red and yellow, respectively.
  • ...and 8 more figures