On Road Trial Claims BEV Power Pack Life can be Predicted

A group of engineers and academics have teamed up in a bid to accurately predict the life of BEV battery packs.

Battery analytics specialist Silver Power System (SPS) has partnered with Imperial College, London EV Company and JSCA, the research and development division of Watt Electric Vehicle Company, on a scheme hoping to solve the issue of battery lifespans that can vary widely from vehicle-to-vehicle. The group’s real-time electrical digital twin operating platform (REDTOP) project has created and trialed ‘digital twins’ of EV batteries.

The value to carmakers would come from having a good idea how long every battery pack will have a working first life. That’s because predicting lifespan has been virtually impossible. While digital models of EV batteries have been created, they have lacked accurate real-world data to back them up. What’s more, not all batteries are born equal and not all batteries are treated equally throughout their life, degrading at different rates, subject to different drivers and charging routines, further underlining the need for real-world data to be combined with machine-learning based predictive technology.

Now real-world data has been fed into the digital twin technology from a trial, run since January in the UK, where about 50 LEVC TX electric taxis and a new EV sports car from the Watt EV Company have collectively travelled nearly 311,000 miles as part of the program.

This crucial data has been analyzed by SPS’s machine learning-powered platform EV-OPS, and together with Imperial College’s battery researchers, the digital twins of working EV batteries have been created, giving a real-world view of real-time battery performance and state-of-health and also the potential to enable these highly sophisticated battery models to predict battery lifespan.

“This really is the holy grail,” explained Pete Bishop, CTO of Silver Power Systems. “Understanding how an electric vehicle’s battery is performing right now – and predicting how it will perform over the coming years – is absolutely critical for many sectors. But to date there has been a lack of data and predictive modelling has been largely lab-based.

“By combining a robust real-world trial with our EV-OPS machine-learning analytics capability through the REDTOP program, we have not only been able to unlock an unprecedented view of real-time battery performance and state-of-health but also create the world’s most advanced digital twin enabling prediction of battery future life.”

— Paul Myles is a seasoned automotive journalist based in Europe. Follow him on Twitter @Paulmyles_

Leave a comment

Your email address will not be published. Required fields are marked *