Obstacle detection an collision avoidance for UAVs utilizing low-cost sensors

Hamish Krippner

Abstract


The presence of Unmanned Aerial Vehicles (UAVs) have rapidly expanded due to the development in technology that has assisted various industries in data acquisition, 3D mapping, and monitoring for search and rescue operations. The increased presence of UAVs is a cause for concern, as it places considerable risk on the safety of other aircraft, infrastructure and the population. This report aims to examine an innovative solution to obstacle detection and avoidance for UAVs and describes the process to develop an effective obstacle detection and avoidance system, utilizing a closed loop design structure for autonomous functionality. The methodology implemented for the open loop obstacle detection system integrates a combination of hardware and software elements which is able to detect dynamic and static obstacles, with a detection rate of 98% at five meters for various sized obstacles. Experimental and numerical results for the closed loop Collision Avoidance System (CAS) demonstrate the UAVs ability to autonomously detect static and dynamic obstacles using an Ultrasonic Sonar (US) sensor and evade obstacles using a guided loop function. All tests showed a collision free path, including adapting to the environment where numerous obstacles were present, demonstrating the effectiveness in a real world application and accuracy of the CAS.

Full Text:

PDF