SafeRide Technologies, a provider of AI-based vehicle health management (VHM), data analytics, and cybersecurity solutions unveiled today vInsight—an AI-based VHM development platform for OEMs, Tier 1 suppliers, aftermarket telematics vendors, and fleets. The vInsight platform includes a VHM development tool and inference engines targeted for in-vehicle and remote AI-based vehicle health monitoring.

vInsight Developer is a tool that enables customers to develop, train, optimize, test, and deploy state-of-the-art deep learning algorithms for VHM using a rich library of production grade models and design templates for major vehicle systems such as engine, transmission, emission, braking, battery management, and more. vInsight Developer can generate compressed and optimized models for embedded deployment onboard vehicles.

vInsight Edge is an embedded VHM runtime engine that is designed for gateway modules, domain controllers, and telematics modules. vInsight Cloud is a VHM runtime engine that supports legacy vehicles and aftermarket deployment. These runtime engines enable real-time inferencing using the trained VHM algorithms created by vInsight Developer.

Existing VHM solutions rely on diagnostic trouble codes (DTCs) and telematics data to detect and predict malfunctions. DTCs can only detect failures after they occur, and they provide limited insights about the problem. This leads to a complex and expensive repair process, and sometimes a wrong diagnosis. Telematics solutions are limited by network bandwidth and access to data; therefore, they fail to address the growing complexity of modern vehicles.

AI technology is vital for delivering advanced VHM capabilities. Nevertheless, domain knowledge and electromechanical expertise are also critical in configuring and tuning the AI algorithms. SafeRide’s vInsight platform is designed with a hybrid approach that combines AI with physics and domain knowledge to deliver optimal performance. vInsight Developer enables customers and partners with vehicle domain expertise to combine their knowledge with SafeRide’s models, templates, and design flows to deliver the required results in a short time, with little or no AI expertise.

vInsight uses SafeRide’s award winning unsupervised and self-supervised deep learning technology. While other AI-based solutions learn from labeled data sets with abnormal health conditions, vInsight can learn from unlabeled data sets with only normal health conditions. This dramatically simplifies the data collection and training processes and enables VHM applications that were not possible before.