Offloading Functions of Autonomous Mobile Robots to a Cloud

Abstract

Autonomous Mobile Robots (AMR) are used in intralogistics to automate internal transport processes. In addition to the increasing number of robots in the system, the demands on the performance of AMR are increasing due to more complex use cases. At the same time, cloud computing is a key technology for handling data- and compute-intensive processes on demand. This work explores the possibility of offloading AMR functions to the cloud and taking advantage of both technologies. To this end, AMR functions are defined and differentiated in a flowchart. After identifying offloading-relevant criteria, the potential of offloading these functions to the cloud is examined using the Analytical Hierarchy Process method. The results show that cross-AMR functions such as traffic control, task assignment, task planning, mapping and path planning have a high potential for offloading to the cloud. On the other hand, AMR functions that require a fast and reliable response time or are safety relevant should be executed on the AMR itself. However, these results underscore that AMRs and the cloud can be combined to great advantage.