Spark EV Technology: accurate EV range prediction
Spark EV has a solution to ease range anxiety
Imagine that you are on the way to an important meeting in your electric vehicle (EV) and the onboard range estimation predicts that you should get there with 10 per cent battery life to spare. Do you press on to get there or do you stop to recharge? To make this decision you have to consider how long will you have to wait, will there be a working charging point nearby and how much will this cost?
Anxiety around charge and range for EVs is perhaps the greatest psychological barrier to mass adoption, and it relates not just to the outright range offered by a vehicle, but the degree of confidence that we can place in these figures.
Calculating these figures is far from a straightforward exercise, even before you consider external factors such as ambient temperature and traffic density. In fact, such is the number of rapidly changing variables involved, that it becomes a prime candidate for machine learning algorithms, which are designed to adapt to changes in seasonality, vehicle performance, battery health and driver behaviour over time.
Big data
Cambridge-based Spark EV Technology has harnessed the power of machine learning for exactly this purpose. The company has created a range prediction system that combines vehicle data, driver behaviour and route information to deliver state of charge estimates that are much more accurate than existing solutions on the market today.
It’s hoped that this technology will help to put drivers and riders at ease, whether that’s commuters or commercial fleets. What’s more, it has the potential to reduce unnecessary visits to public charging points, freeing up infrastructure for other EV users, reducing costs of overcharging and generating more accurate data for vehicle manufacturers and fleet managers.
Spark EV Technology was awarded £104,500 of grant funding in 2020 as part of the Technology Developer Accelerator Programme (TDAP) from the Advanced Propulsion Centre (APC). The programme also gave the company access to networking opportunities, detailed advice on how to commercialise its technology and platforms to showcase it.
“We wanted to mature our technology, but also to expand our access to different people within the industry, new customers and new partners,” comments Spark EV Technology CEO, Justin Ott. “TDAP has helped us to develop the robustness of our technology, enhancing our technology readiness level and expanding the operating systems which can integrate our product.”
Concept car
As result of its work on the TDAP project, the company was able to integrate its range prediction system into the Aura EV – a concept car funded by the Niche Vehicle Network. TDAP also helped to establish new relationships, both within the passenger car market and beyond, working with EV manufacturers of all sizes.
“One piece of feedback that we had early on in the programme was, ‘you have this prediction technology for four wheels, why don’t you look at two?’” comments Ott. “We were introduced to an electric bike manufacturer and were able to develop, test and prove our technology with two-wheeled customers, which has led to a new commercial partnership with a major Tier 1 [direct to manufacturer] supplier.”
Discussions are ongoing with a number of other potential customers, and it’s hoped that the technology successfully demonstrated on the Technology Developer Accelerator Programme will allow Spark EV Technology to secure further funding in a Series A investment round later this year.
