Wunelli develops Driver Signature
Wunelli, a LexisNexis Risk Solutions company, has developed Driver Signature analytics for smartphone telematics apps to identify a driver after just 200 miles of journeys.
Having accumulated billions of miles of driving behaviour data and hundreds of thousands of related claims, Driver Signature uses machine learning to identify when a policyholder is driving and offers a 15% improvement in accuracy compared to the rules-based approach commonly used.
Driver Signature is identified using a range of metrics based on where, when and how motorists drive – characteristics that are individual to each driver. After collecting a certain number of trips meeting the minimum driving distance, a set of driving pattern attributes will be developed and a Driver Signature analytics model applied to classify the trips as either a driver trip or a passenger trip. This is then used as the basis for scoring all future journeys.
There are recommended default settings, but an insurer will be able to choose customised parameters, including the minimum number of trips and minimum distance of trips, as well as the classification threshold. There is no customer interaction needed during this process, although insurers will have the option to manually override the model classification results in the app.
Paul Stacy, founding director of Wunelli said, ‘The advances we have made in Driver Signature analytics represent a significant step forward for telematics apps. This improves both the customer experience and removes the risk of fraud – that is, policyholders claiming to be driving when they are not and vice versa. This development should bring insurers and brokers a new level of confidence in the use of smartphone apps as a means to collect driving behaviour data to support telematics based insurance.’
