As a consultant and industry advisor to Auburn University's Transportation Institute, I love it when a major trucking fleet takes out a full page ad in an industry publication which, in my mind, asked a group of Auburn students and me for help. Here is what the ad says:
We're making it impossible to break down in the middle of nowhere
With remote diagnostics, we not only know if there's a problem with your engine, we know from 800 miles away. By monitoring vehicle data, we can help you make reliable, informed decisions, getting you back on the road quickly – and keeping you there. It's how we deliver confidence.
What caught my attention was the specific reference to the engine. This, of course, is a great improvement over the past. But, of all the headaches with trucks the engine is near the bottom of the list. Remote diagnostics on engines is great, but remote diagnostics on everything that could fail – enabled with artificial intelligence (AI) – would be better.
Tires and wheels are near the top of the problem list for commercial vehicles. I happen to know for a fact a group of Auburn University engineering students are working on getting wheel end data that can be monitored by an AI and Machine Learning (ML) system. An informed decision to stop the vehicle before a tire becomes a road gator (or the entire wheel and tire leave the vehicle) can be made by monitoring heat and vibration at the wheel end. This team is working on a shoe string budget. They would have made the Skunk Works proud.
Using the Tangerine Innovation AI and ML system, senior design students are monitoring a set of shock absorbers. AI will send an alert when one shock is not the same temperature as the others. If one shock doesn't heat up like the other three, for example, AI would report a leaking shock. What if out of the four shocks, one is getting extremely hot compared to the other three? That tiny bit of data might be an indication of loose lug nuts or a separating tire. Machine Learning is a great system for monitoring and comparing components and predicting failures before they occur.
