Data received from a vehicle-mounted telematics device comprises accelerometer data and corresponding vehicle location data. The data (processed to be indicative of a roughness of said road) is applied to a model relating roughness data to pothole occurrence to derive a pothole forecast. The model is generated from stored roughness data derived from similarly processed accelerometer data sampled from roads with known pothole occurrence. The processing may comprise determining vertical acceleration size and frequency. The data may be filtered to remove periodic or background accelerations indicating road surface of traffic calming features. Data inputs to the model may comprise combined vertical acceleration data per unit length of road and vehicle speed and outputs may comprise an expected rate of pothole occurrence per unit length. Also disclosed is receiving a pothole forecast for a road at a vehicle and outputting an indication to a driver where it indicates a probability of pothole occurrence to exceed a predetermined value or using it in a route planner.
Pothole monitoring
2020-09-23
Patent
Elektronische Ressource
Englisch