Development of a cloud-enhanced investigative mobile robot

摘要

In recent times the cloud based applications for robotics has become an emerging topic of discussion. Both individual as well as networked robotic operation can be benefitted with the support of cloud computing paradigm. The different applications of the cloud enhance its integration with various aspects of robotic operation. However, it depends on the scope of the applications as well as the setup of the robotic system. This has led to our work in this paper that includes a hardware implementation of a robot with cloud-aided applications of face detection and identification for the purpose of interaction and investigation. We explain our face recognition algorithm that we have implemented as well as our methods of using it for context-aware offloading to the cloud. In addition to that, we present an autonomous platform switching algorithm that is network and context aware in terms of task allocation between resources. Our practical implementation and demonstrations analyse the impact of cloud in the performance of this application. From the results, we can suggest that cloud-based approach is beneficial for performance and longevity of the system.