The objective of this project is to validate and develop a new concept of road survey and management based on big data collected from standard sensors installed on modern vehicles. The project will deliver a decision-making tool accessible by the road authorities and municipalities which will provide reliable and live monitoring of the road conditions. Limits of standard road surveys and relative costs will be overtaken and the final users will be able to optimize maintenance strategies and guaranty safety conditions. Nowadays it is possible to access georeferenced cars data on which are hidden road physical conditions. By using advanced machine learning, it will be possible to develop models, which measure road characteristics by using data from conventional car sensors. In fact, machine learning allows computer systems to learn directly from examples, data, and experience. Machine learning systems can carry out complex processes by learning from data, rather than following pre-programmed rules. In this specific case, machine learning will help to translate the vast amount of data collected by the cars into road properties.
GreenMobility, with its electrical cars fleet, represents an ideal platform for a consistent data collection where bias due to difference in vehicles and sensors are overtaken. The municipality of Copenhagen is the perfect sand box because offers a wide range of infrastructure types which are fundamental during models’ development and validation driven by machine learning. It is expected that this new concept of road scan and management will represent the beginning of a new era and business. Rental car, bus and insurance companies will have the interest on selling or using data collected every day for free on the road network to support road owners and authorities.