Abstract

With the growth of wireless networks and Internet resources, cloud robotics is becoming more and more popular. A prominent innovation that has gained traction is computation offloading, which gives robots more computational power by using cloud resources and parallel processing power. Nevertheless, because there are so many variables affecting performance, research on how effective computing offloading is in cloud robotics is still underway. Our research proposes a distributed cloud robotic architecture that uses Kafka middleware as a message broker to offload compute in order to overcome this problem. We evaluate the performance of our solution using image processing techniques through experimental validation. Our tests show a significant decrease in the robot's CPU burden, but an anticipated rise in the robot's communication latency.