A hybrid approach to automotive AI
Truck fleets have long used sensors to monitor parameters such as fuel use and driver behavior. Intangles Labs in India is ratcheting the monitoring function up a notch by processing monitored data via its custom-built neural nets combined with conventional rules-based approaches. The sensing used to gather data for processing includes custom modules that connect to the vehicle data port – an example of which is held here by Intangles’ Aniket Didolkar – but the system also works with ordinary sensors found on engines and chassis. Typical information gained this way include mileage for routes based on load and road conditions, identification of poorly performing drivers, insights into idling, hardbraking, overspeeding, and transmission use over distances traveled; real-time visualization of vehicle engine parameters, and so forth.
NEXT PAGE: Just the thing for the absent-minded
Hypervisor for automobile is a smart idea if the overhead proves insignificant. Someday automakers can realize the security benefits of hypervisor.