Design of a new path planning and obstacle detection system for distributed mobile robots

In industry, mobile robots are generally used for load transportation between workstations in the factory. If these loads are carried by forklifts using conventional methods, errors and accidents may occur due to the operator. Furthermore, the movement limitations of the forklifts also create a limited working area during the transfer of loads. The use of autonomous mobile robots in industrial areas is important in reducing these errors and accidents. One of the common uses of mobile robots that are sensitive to environmental factors and can work autonomously is the warehouses where the stocking service is provided. For these robots, which are usually used as robot fleets is carried out in many academic studies along with commercial developments. Solving localization and positioning problems is an important part of these studies. The fact that robots move safely along the route with minimal power consumption while performing the specified tasks constitutes the basis of academic and commercial studies. In this thesis, a study based on probabilistic particle filter algorithms has been carried out on the path planning of mobile robots with distributed mobility. In the study, firstly, 2D map of the field where robots are located have been mapping by SLAM and Occupancy Grid Mapping methods. Using these methods, the multiple mobile robots within the map were accessed from a designated starting point to the destination using the Adaptive Monte Carlo Localization algorithm. The steps to reach the destination on the map known by the robots and the navigation parameter are controlled in real time via a master station. These processes are performed via SSH protocol over a wireless network. The dynamic obstacle avoidance of the robots is provided by realtime images taken from the kinect camera while route on the map. For this local path planning, Dynamic Window Algorithm was used. In the thesis, a system design and prototype which can perform many functions such as mapping, positioning and localization using modern algorithms and equipments have been achieved. In this respect, the thesis will contribute for R & D studies of mobile robots of different structures and features which are needed in many sectors those Industry 4.0 applications. Keywords: Autonomous robot, Adaptive Monte Carlo Localization, path planning, obstacle avoidance, robot operating system