A Study on Real-Time Data Transmission of Cloud Controlled Mobile Robots

Mobile robots have very limited computational and storage resources. However, they must achieve tasks that are too heavy to be processed by their processing units. Additionally, other mobile robot applications require sharing information or building centralized maps to achieve cooperative tasks. Cloud computing is a good option for offloading heavy computation tasks necessary for robot navigation. The computation nodes can store the processed data in unlimited cloud storage. However, this solution flawed with several issues that include real-time transmission of data between mobile robots and the cloud server. This paper demonstrates the mobile robot control system using a cloud server while considering the size of the data transmitted between the mobile robot and the cloud server. This is achieved by compressing the sensor data and converting them to topological maps. Experiments demonstrate great efficiency enhancement of data transmission under a limited communication bandwidth, consequently increasing the scalability of the system.