Accelerating EV adoption – a better range prediction helps to create credibility and leads to customer trust and greater acceptance

A project between Bosch and Spark EV Technology

The consideration of buying Electric Vehicles (EVs) has notice- ably increased over the past few years, as consumers acknowledge their numerous benefits. Despite the recent rise in EV sales, concerns regarding batteries, charging and driving range continue to prevent a large-scale consumer pull.

Range anxiety is still one of the biggest hurdles for widespread EV adoption. Users rely on the dashboard’s ‘remaining distance’ in km to judge if they can safely reach their destination or if there is a need to charge. This ‘remaining distance’ often turns out to be inaccurate, since it does not adequately take into ac- count all range-affecting parameters of the upcoming trip, such as road slope, outdoor temperature, driving style and traffic.

To tackle the challenge of inaccurate predictions and letting drivers know how far they can drive before it runs out of battery, German Tier 1 Bosch has collaborated with Spark EV Technology, a UK based startup that develops and supplies intelligent range prediction software. The aim of the project was to provide best-in-class, highly accurate range predictions giving drivers the confidence to use their EV to the fullest, accelerating electric vehicle ownership and improving brand reputation for Bosch’s renowned customers.

The Project 

The project was divided into data collection and data analysis. During the data collection phase, Bosch used a test app provided by Spark EV to log trips with an electric vehicle. A vehicle with a 41 kWh battery was chosen as a test vehicle and the driver was given the freedom to undertake journeys of their choice on different routes in the Stuttgart region of Germany over a period of two months.

Additionally, Bosch utilised an On-board Diagnostic (OBD) dongle to retrieve detailed data from the vehicle. In total, around 30 trips were collected. Due to the COVID-19-induced lockdown, those trips happened to be mainly short trips, around 16 km on average,  very often to and from the same location. It was agreed that the journey data would be recorded before and after each trip into a mobile app provided by Spark EV – allowing for simple data collection and would enable the predictions to be compared in real time.

This method of data collection was to ensure the trials would represent true and fair real-world testing of Spark’s range prediction technology while comparing it to the OEM’s range prediction system currently embedded in the vehicle.

The collaboration with Spark EV was a success proving that the indicated remaining distance provided by the manufacturer’s dashboard is highly inaccurate. In addition Spark’s technology managed to consistently increase range prediction accuracy for trips with given parameters (start, destination, weather, traffic). For instance, the average prediction error by the test vehicle was 10 km, versus

1.4 km prediction error provided by Spark EV. Given that the average trip length was 16.4 km, this is a significant improvement.

Future Outlook

Going forward Bosch plans to further investigate the applicability of the results and if successful, Bosch and Spark will explore commercial opportunities to benefit their OEM customers.

EDITOR: TIMO LINNEMANN

CREDIT: STARTUP AUTOBAHN powered by Plug and Play