An autonomous robotic system for intralogistics assisted by distributed smart camera network for navigation

Abstract

Versatile transport systems with a swarm of mobile robots used in intralogistics have a great advantage in adaptability and flexibility. In this work, a concept of the low-cost autonomous transport system with distributed smart cameras is researched to overcome some problems for adapting to future requirements of a smart factory, e.g. high capital costs, and lack of timely and accurate environmental knowledge. Computational resources are distributed to infrastructure sensing agents to process local optical data from cameras. The transport agents navigate by flexibly collecting the desired spatial and temporal environmental information from the smart cameras based on different levels of interests. In addition, a temporal occupancy map update mechanism and an observation correction model are presented to improve the system performance. The concept of our proposed system is demonstrated in a designed workspace, and the experiment results show that the presented concept is feasible.