A control middleware for cloud robotics

Future robot applications will require cloud computing. Using an architecture of cloud robotics would increase performance over hardware constraints (e.g., processors, memory, and power). In this paper, we propose a control middleware for cloud robotics, which not only can receive the robots' requests and allocate the cloud resources based on the robots demand, but also can ensure the connection between the robots and the cloud more stability. And then, we use an autonomous, humanoid robot to test the first stage of the control middleware. Effectiveness of this control middleware is demonstrated by experiments.