VTNZ is New Zealand’s leading vehicle inspection company. Together with two partners, it developed a prototype for a predictive-maintenance solution, based on comprehensive vehicle inspection data generated in over 150 locations. Once released, the solution will help car owners identify mechanical problems before they happen. What happens if you combine the knowledge of VTNZ’s experienced inspectors with state-of-the-art computer technology? Well, what you get is a machine-learn-ing solution that can look at large amounts of data quickly and predict what is likely to go wrong with a specific car model, based on parameters such as type of vehicle, location, and age. Naturally, VTNZ’s prototype focuses on areas that might be subject to failure based on Government inspection requirements, such as lights, brakes, chassis, and oil leaks. However, additional data from onboard computers could be included to expand the range of predictions. The computer continuously learns and adjusts its prediction method based on new information from inspection staff and inspections recorded digitally every day, making it much more sophisticated than any kind of statistical analysis. Currently, the solution is around 90 percent accurate in its predictions and shows a lot of promise. 

­