Real Time Tire Wear Detection Using Intelligent Tires

Abstract

Abstract Tires, the only vehicle component in contact with the road, are crucial in determining vehicle dynamics. However, as the tires travel substantially, different wear patterns such as feather, spotty, and camber wear are introduced in different parts of the tire’s tread due to contact with the road surface, improper tire inflation pressure, and tire alignment issues. This tire wear phenomenon affects vehicle performance and is detrimental to vehicle-road safety. This research has developed a framework for tire wear classification based on the “Intelligent Tires” concept. This intelligent tire comprises tri-axis accelerometers mounted on the tire’s inner liner for data corresponding to the tire-road contact interface. A vehicle was instrumented with different proprioceptive and exteroceptive sensors for testing and data collection. Tires with different tread depths have been tested under various test conditions of velocity, tire inflation pressure, road surface, and tire normal load. Multiple signal processing techniques in time, frequency, and spatial domains have been leveraged for feature extraction. These features serve as inputs to machine learning based classification algorithms to distinguish between normal and worn-out tires. Thus, the developed system capitalizes on intelligent tire data to predict the state of the tire as normal or worn-out, with real-time capabilities.