Road related telematics encompasses a combination of road transportation, in-vehicle electronics and telecommunications. It is continuously evolving in terms of complexity and diversity, and contains high volumes of sensor data. Telematics-based technology is mainly used for the recovery of hijacked/stolen vehicles, insurance purposes and vehicle fleet monitoring and management. In this paper, it is demonstrated that telematics-based technology may offer capabilities in terms of predicting road condition. This is based on a project where some simple statistical analysis techniques were used to interpret existing data originating from standard vehicle telematics units. The paper mainly evaluates the z-direction (up / down) acceleration for a few different vehicle types and x-direction (lateral), y-direction (longitudinal) and z-direction acceleration for the identification of road anomalies/defects. Roads are currently monitored using Class 1 profilometers, which provides a detailed but relatively costly indication of road condition. The use of vehicle telematics may provide a more cost-effective solution to monitor a wider road network and on a continuous basis, albeit on a Class 3 profilometer level. Telematics units are currently installed in numerous vehicle types throughout South Africa, and this paper focuses on achieving a reliable Class 3, real-time response-type measurement which can be used as a screening device and which is able to accurately produce an indication of road roughness, as well as major road distresses. This may assist in ensuring that agencies without direct access to funding for Class 1 road condition data may be able to obtain an indication of their road network conditions. It also contributes towards the safety and comfort of road users.
Paper presented at the 34th Annual Southern African Transport Conference 6-9 July 2015 "Working Together to Deliver - Sakha Sonke", CSIR International Convention Centre, Pretoria, South Africa.