Modular Navigation Strategy for Autonomous Robot

Abstract

The evolution and democratization of different technologies, such as Wireless communication, size and calculating power of computers, sensors capacities, image treatments, plug and play mechanical components with its electronic command during the last ten years, ease nowadays designing mobile robots and their control algorithms. According to the robotics literature, different methods are available to perform the obstacle avoidance task: edge detection, cell decomposition, map building, artificial potential fieldand so on. This paper presents a project overview of designing a universal command and control architecture dedicated to wheel driven autonomous and semi-autonomous robots. Two nonlinear control techniques for trajectory tracking and obstacle avoidance have been developed and implemented. The platform should be able to move towards the goal while automatically avoiding obstacles. To accomplish this task and navigate to the desired targets while reasoning intuitively thanks to the fusion of its sensors data, the robot could rely on a wireless human assistance with a mobile interface, when it's necessary. Hazardous structured indoor and specific outdoor interventions are the main use of this robot autopilot.