Thu 12 Oct 17
Our computer scientists have helped a Cambridge company develop journey prediction software to make electric vehicles more efficient between charges.
Dr Michael Fairbank and Dr Daniel Karapetyan of the School of Computer Science and Electronic Engineering, worked with Spark EV Technology to develop algorithms to help electric vehicles travel further between charges, and eliminate the stress of drivers not knowing whether they have enough power to complete their journey.
The software gathers live data on factors such as weather conditions, the volume of traffic, tyre wear and driver behaviour to predict how much energy is needed for each journey. Spark uses machine learning to compare the prediction against the actual energy requirements to increase the accuracy for each future journey.
Dr Michael Fairbank, explained: “Often electric car drivers travel shorter distances per charge than they could to avoid running out of power. This data can extend the vehicle’s range by an average of 20% between charges, so it is particularly good for companies running a fleet of taxis or delivery vehicles, as it reduces driver downtime.”
Justin Ott, Chief Executive Officer Spark Technology said: “Companies around the world are modernising their fleets right now and are looking at the potential of electric vehicles. Spark Technology gives fleet managers reassurance that vehicles will complete journeys while providing vital data on how vehicles and their drivers perform in different conditions. More electric vehicles on the roads mean fewer carbon emissions, leading to better air quality for all.”
The project was recently awarded investment from the Low Carbon Innovation Fund overseen by the University of East Anglia. The fund provides equity finance for small and medium sized enterprises (SME's) in the East of England that are contributing to the low carbon economy.