Towards a General and Complete Social Assistive Robotic Architecture

Social and service robots are designed to achieve complex tasks that are increasing daily. Cloud resources enhance service robots with strong computing capabilities and higher data storage centers. One of the major drawbacks of the cloud robotics model is the network latency that edge computing addresses. Considering the capabilities that a service robot must have to achieve somehow human-like capabilities, we propose a hybrid service-oriented fog robotic architecture. The proposed architecture follows the fog robotics approach to distribute the computations through the cloud, edge, and robot in a clear way. Robotic control systems are significantly varied based on the required tasks, environment, robot specifications, user and business needs, or programming tools that were used. The paper introduces several concepts to facilitate the designing or developing process of the robotic system based on the desired robot’s type and functions. Furthermore, automated planning, one of the most important factors to an intelligent and autonomous robot, has been evaluated based on several scenarios. These cases address the computational and the planning aspects through either the Edge or the Cloud server.