Jet Infosystems developed for Kontrol leasing the solution based on Machine Learning

2018

On October 17, 2018 the Jet Infosystems company reported that it developed for KONTROL leasing the solution based on Machine Learning (ML) which in real time registers deviations in behavior of drivers. According to the company, in the automatic mode cases of dangerous driving, potential stealings and breaches of agreement of leasing are traced. The KONTROL leasing company pays special attention to safety of use of objects of leasing, control of a status and wear of the cars, and this solution allows to reduce the number of accidents and financial risks.

The mathematical model of the solution in 2018 studied at the data obtained from cars of the company. According to the results of the analysis of a large number of parameters in work of cars and actions of drivers it was succeeded to create basic models of standard behavior in different conditions. A system with an accuracy distinguishes standard and non-standard actions, is able to distinguish temporary changes in behavior from characteristic style of driving.

The telematic device is connected to the car and conducts collecting of signals with all main systems of the vehicle. Data of the tachometer, accelerometer, clicking and release of a pedal of a brake, opening and closing of doors, inclusion of the mode of the parking and other parameters are registered. The solution reveals deviations, uncharacteristic for specific drivers, in behavior and sends the notification to the staff of the company.

It is obvious that manually it is impossible to control work of all leasing park. Machine learning technologies in a root change a situation. A system is capable to select independently cases of atypical or suspicious behavior which increase risks of our company. Reacting to such cases, we often prevent emergence of adverse situations and we avoid possible losses. Sergey Zherdev, director of information technology of KONTROL leasing company

The automated analysis and control of actions of drivers helps to prevent cases of stealing and unauthorized use of transport, to reduce wear and indicators of accident rate. In addition there is an opportunity for collecting and transfer to insurance companies of the fair drivers given about accident-free driving.

The project shows that at the correct use of Machine Learning technology can improve not only business indicators of the company. In this case they help to increase safety of driving, so, and safety of passengers – users of services of transportation. It is demonstration of potentialities of artificial intelligence. Vladimir Molodykh, director of development and deployment of Jet Infosystems company