Cloud-Powered Digital Twins: Is It Reality?

Abstract

The flexibility of future production systems envisioned by Industry 4.0 requires safe but efficient Human-Robot Collaboration (HRC). An important enabler of HRC is a sophisticated collision avoidance mechanism which can detect objects and potential collision events and as a response, it calculates detour trajectories avoiding physical contacts. Digital twins provide a novel way to test the impact of different control decisions in a simulated virtual environment even in parallel. The required computational power can be provided by cloud platforms but at the cost of higher delay and jitter. Moreover, clouds bring a versatile set of novel techniques easing the life of both developers and operators. Can digital twins exploit the benefits of these concepts? Can the robots tolerate the delay characteristics coming with the cloud platforms? In this paper, we answer these questions by building on public and private cloud solutions providing different techniques for parallel computation. Our contribution is threefold. First, we introduce a measurement methodology to characterize different approaches in terms of latency. Second, a real HRC use-case is elaborated and a relevant KPI is defined. Third, we evaluate the pros/cons of different solutions and their impact on the performance.