IoT-robotics for collaborative sweep coverage

The combination of Task Scheduling (TS) approaches in Multi-Agent Systems (MAS) and Path Finding (PF) can produce a wide range of solutions in several application contexts, such of logistics, sweeping and cleaning of large areas, and surveillance missions by Unmanned Vehicles. This paper presents a task assignment method for multi-agents-based collaborative sweep covering, where it is relevant to follow the optimal route in order to maximize the expected results with the available resources and constraints. The designed solution uses a smart planner, which computes optimal routes, and a centralized scheduler that assigns tasks to unmanned robots according to different priority queues. The prototype implementation integrates off-the-shelf IoT technologies to drive a simple robot in a controlled environment. Image processing technologies are used either to estimate in advance the expected reward for the planned route and afterward to get a feedback about the task execution. • The paper presents an optimal path planning and Job assignment method for multi-agent-collaborative sweep covering. The method takes into account available resources and environmental constraints to coordinate IoT devices. • A smart planner is used to calculate the best routes using an algorithm based on a posteriori probability, while a centralized scheduler distributes tasks to unmanned robots, with the goal of improving efficiency and task completion. • In the paper we combine planning algorithms with Computer Vision algorithms. The planner, utilizes feedback from IoT devices elaborated through a computer vision algorithm. This feedback mechanism ensures dynamic re-planning and optimization of tasks based on real-time data.