Invers adds AI-based damage detection to digital shared mobility solution

• Invers combines its reliable shared mobility solution with precise, AI-based damage detection by carValoo
• The new solution addresses a major obstacle for operators on their way towards profitability
• Based on reliable sensor data, AI-based algorithms detect even minor damages and other safety critical events in real time, enabling efficient vehicle handovers and optimized claims processing for providers of car sharing, car subscriptions and rental

Siegen, February 28^th, 2023 – Invers is further expanding its proven solution for shared mobility by carValoo’s AI-based damage detection: the new solution offers digitized and automated damage detection based on data from highly sensitive motion sensors within the car. These data are analysed and processed by artificial intelligence to detect damages precisely and reliably. The solution immediately contributes to business success as it allows operators to reduce costs by improving damage management, increasing successful damage claims and completely digitizing the rental process. Also, it is easy to install and quickly set up.

Only one in eight carsharing customers who cause damage to a sharing vehicle report it to the operator. As a result of this daily occurrence, carsharing operators lose up to 10 percent of revenue because they do not have sufficient information to claim damages. Enabling them to realize this significant share of revenue lets them be more profitable and use these profits to increase investment in their services and spread shared mobility even further. Also, damages that are not reported mean that the next customer faces safety risks if their car’s roadworthiness is impaired. In addition, operators lose out on the resale value of their cars since they usually do not carry out regular checks, and so do not detect minor and medium damage until it’s time to sell their vehicles. Damage detection is also a challenge for traditional car rental companies who do carry out regular, resource-intensive visual inspections that are unnecessary in approximately 90 percent of vehicle returns and usually only detect externally visible damage.

The new solution from Invers and carValoo provides detailed information on vehicle damage in real time and enables comprehensive, automated damage management. “Invers has been innovating in car sharing technology for 30 years”, says Alexander Kirn, CEO of Invers. “This innovation is another example of how we use technology to benefit all stakeholders in the mobility ecosystem to ultimately make the use of shared mobility more convenient than car ownership.” The partners have combined their technologies in such a way that the solution can be installed quickly and easily via plug&play. “Providing carValoo’s proven AI damage detection through Invers sharing technology makes the service even more efficient and convenient for customers”, says Tom Althoff, Managing Director at carValoo.

Highly sensitive acceleration and gyro sensors in the new solution capture various data from a vehicle in daily use. The motion sensors thus generate 72.000 data points per minute per car which sums up to 103.7 billion data points per day per car. Not all of them are related to damages and are relevant for damage detection. For example, closing a door of a parked car or driving through a pothole creates sensor data points similar to a parking accident. The solution’s intelligent algorithm already filters and compresses a large part of this raw data already on the sensor box within the car, so that less than ten percent of the data points still need to be transferred to the cloud. The data points that arrive in the cloud are then further analyzed in detail there by an AI algorithm which identifies patterns and anomalies. The AI then assigns them in real-time to specific events that, like a parking accident, have actually led to damage to a particular component. At the same time, it recognizes if it was only a door being closed or a pothole being driven over, resulting in no damage and making a report unnecessary.

In addition, the solution is “learning” from big data sets and is constantly improving: since its launch three years ago, damage reports have been compared with actual damage, so that the artificial intelligence continuously trains and optimizes its accuracy. The system reliably identifies even minor damages in over 95 percent of cases and locates them correctly on the vehicle. Alongside the fully digital solution including sensors, an API and the backend, this precision is a key unique selling point. It also allows damage that is not detected in normal visual inspections to be identified. Fleet operators gain very precise insights into the condition of their vehicles, can automate their damage management and claim compensation or deductibles.