DYNAMIC TASK ALLOCATION AND PATH PLANNING IN SINGLE-AGENT ROBOT WITH CLOUD COMPUTING: A SURVEY
Hadeel Khudhur Al-Ahmedy, Saif Alabachi, Ibrahim Adel Ibrahim · Kufa Journal of Engineering · 2025
Dynamic task allocation is an essential aspect of robotics especially when achieving frequent mobility or running sensitive tasks that must be done with extreme precision and efficiency. Several objectives such as achieving high accuracy in task execution, accurate path mapping, successful target reach, precise trajectory prediction and optimized control which are increasingly facilitated by the integration of cloud computing. Traditionally, multi-agent robotic systems have been utilized where the overall path is divided into smaller partitioned tasks, then each robot is assigned a specific task and the entire system is optimized using variety of hybrid optimization algorithms. Neural networks are used for automated prediction, while cell decomposition methods help in task allocation and path planning. In this study, we introduce an effective method by using a single-agent robot with cloud computing. We highlight the advantages of this method over traditional multi-agent systems particularly in scenarios requiring comprehensive path information. Our approach leverages optimization techniques, cell decomposition and Elman neural network to streamline task allocation and execution. The results show that the proposed approach achieved accuracy about 92% with task completion time around 8 minutes. Although the multi-agent robot performed higher accuracy nearly to 95% and less time about 5 minutes, the single-agent robot significantly reduced a computational load and energy consumption. This approach consumes 45 Watt-hour (Wh) compared to the multi-agent strategy which consumes more energy 60 (Wh). In conclusion, this review article improved the efficiency and effectiveness of dynamic task allocation and path planning in robotics and offered a responsive solution for complex environments