Collaborative Perception in Multi-Robot Systems: Case Studies in Household Cleaning and Warehouse Operations

TL;DR

Collaborative Perception aims to overcome limited sensing of single robots by fusing data from onboard and environment sensors into a unified representation via a central processing server. The paper demonstrates CP through two case studies: a home-cleaning team coordinated by overhead cameras and AMRs navigating a warehouse with aisle-mounted lidars, showing improved task efficiency, coverage, and safety. The approach relies on a centralized perception server that aggregates data and supports workload allocation, while path planning in the warehouse uses $ARA^*$ to compute efficient routes. Findings indicate CP reduces travel distance, planning and replan overhead, and increases throughput, with implications for broader domains such as healthcare, agriculture, and public safety; future work will address bandwidth, latency, and data security to enable real-world deployment.

Abstract

This paper explores the paradigm of Collaborative Perception (CP), where multiple robots and sensors in the environment share and integrate sensor data to construct a comprehensive representation of the surroundings. By aggregating data from various sensors and utilizing advanced algorithms, the collaborative perception framework improves task efficiency, coverage, and safety. Two case studies are presented to showcase the benefits of collaborative perception in multi-robot systems. The first case study illustrates the benefits and advantages of using CP for the task of household cleaning with a team of cleaning robots. The second case study performs a comparative analysis of the performance of CP versus Standalone Perception (SP) for Autonomous Mobile Robots operating in a warehouse environment. The case studies validate the effectiveness of CP in enhancing multi-robot coordination, task completion, and overall system performance and its potential to impact operations in other applications as well. Future investigations will focus on optimizing the framework and validating its performance through empirical testing.

Figures (7)

• Figure 1: Pipeline for multi-robot decision making based on collaborative perception.
• Figure 2: View from overhead cameras
• Figure 3: Example output of ML pipeline run on the server
• Figure 4: Cleanable area segmentation results
• Figure 5: Warehouse environment used for analysis