ARRANGEMENT AND METHOD FOR PREDICTING ROAD FRICTION WITHIN A ROAD NETWORK

A system and method are described for predicting road friction using data from a fleet of connected vehicles. Each fleet vehicle includes a communication arrangement for reporting, to a back end system, floating car data sets, including a position of the vehicle, time data and data regarding determined road friction influencing parameters and a determined road friction associated with that position. The data sets are collected and aggregated in a central database, over a predetermined time period. A neural network computer is trained and validated, using aggregated data sets, to create a model to predict future road friction for a specific road network position or segment associated with that specific position. Once trained and validated, the neural network computer, upon receiving the same type of new, up to date, input data, uses the model to predict future road friction for that specific position or segment of the road network.