Evaluation of a Collision Avoidance System at an Underground Mine

Collision avoidance system's design and configuration introduce operational delays, particularly in mines where several mobile machines and workers interact. Following the deployment of a radio-frequency identification collision avoidance system on underground loaders at a platinum mine in South Africa, production decreased by 13.28%. This research study focusses on determining the system's impact on productivity, its constraints, ranging and detection accuracy. By triggering alarms and measuring activation distances for stop, crawl and caution mode the system was assessed on surface and underground in static and dynamic trials. Caution mode was the most accurate and crawl mode the least, rear direction was the safest and front was least safe, the system performed better underground than on surface. Metallic parts of the loader which were in line of measurement, caused tag detection failure in front of the bucket during surface trials, and distortion in distance estimation which influenced productivity. Utilising Received Signal Strength technology rather than Return Time of Flight may improve the system's accuracy because of its even magnetic field distribution in the presence of metallic objects.