An Edge Computing Sizing Tool for Robotic Workloads

Robots of today are equipped with lightweight computing resources used merely to make the robot function. However, proportional advancements in associated data processing and algorithms are needed, given the significant advances in robots’ sensing and programming capabilities and the increasingly complex tasks they must complete. Yet, such advancements require additional hardware resources to function as intended. In many robotic applications, cloud computing is not an option; therefore, edge computing must be embraced. This paper proposes a sizing tool for benchmarking workloads against pre-written tasks to determine optimal edge computing hardware candidates used to deploy said workloads efficiently without wasting resources. Preliminary results show that the right combination of hardware resources has an impact on workload execution.