Range prediction for efficient electric last mile solutions

Precise battery range prediction is crucial for last mile electric vehicles. Discover how our software-based continuous vehicle mass estimation can help.

More and more municipalities in Europe gradually change applicable laws in order to ban vehicles with polluting combustion engines, allowing only modern low-emission or electric vehicles to drive in inner-city areas. Among these places are capitals like Amsterdam, Paris, London, Rome, Brussels, but also other large cities like Stuttgart, Manchester, Lyon, Milan, to name just some of them. Altogether, there are already over 250 so-called low-emission zones in Europe. They share a common goal, namely to improve air quality and to accelerate the switch to zero-emission electric vehicles.

In Paris for instance, there will be a total ban on diesels as early as 2024. As well, Amsterdam will follow with a ban on both diesel and gasoline vehicles from 2030. As the average age of light trucks on the road in EU is currently more than 11 years, impacted fleets and companies will have to deal with this topic in order to sustain their business and their ability to offer services in low-emission zones. The actual result is that last mile delivery fleets need to start their transition to electric utility vehicles right now in many urban areas, in order to increase the share of electric last mile vehicles in their fleets.

July 14th, 2021, the European Union has also proposed an effective ban for new fossilfuel cars from 2035. No doubt that for the electric vehicles’ industry, this represents a tremendous growth potential combined with ever-increasing e-commerce transportation.

A range of electric light trucks for last mile delivery are available on the market. Their typical battery range is somewhere around 150 – 200 miles (roughly 250 – 300 kilometers). The effective battery range of each vehicle depends very much on factors like

• Equipment manufacturer and vehicle model
• Battery health status
• Driving style
• Environmental conditions, and
• Vehicle mass.

General vehicle data are fixed, and many other data can be monitored via telematics solutions (speed, outside temperature, …) but not the loaded mass, which has an huge impact on the actual range. You cannot roll a small delivery truck over a scale at every stop, nor can you type this data manually into its onboard computer.

The issue is not just about the nominal range of an urban utility vehicle rolling off a production plant, but rather the actual result how a truck behaves in your day-to-day business, about your expectations with respect to charging frequency, and most important of all: reliable battery range prediction whatever payload has been picked up at a stock or warehouse. Constant changes in the loaded vehicle mass are inherent to last mile delivery services, but the loaded mass reduces the battery range. This makes range prediction of commercial electric vehicles so difficult. Currently, electric vehicles and electric trucks do not have the ability to measure their own instantaneous mass. The battery range displayed on the instrument panel is based on energy consumption history.

Under constant cost pressure, each delivery company is searching for means to optimize fleet operation, including the number of deliveries with one single charge. If the remaining battery range is not well predicted, there is a risk that a last mile delivery truck runs out of battery during operation. Unplanned breakdowns disorganize the fleet planning, lead to additional costs and customer dissatisfaction. Fleet professionals need to deal with reliable information on remaining battery range.

But how to know precisely, how many miles of range an electric truck has left before the next charging?

COMPREDICT has developed innovative products based on machine learning. Our company’s virtual sensors based on standard CAN bus signals are able to calculate instantaneous vehicle mass with a precision better than +/-5%, after 2-5 minutes city driving. Newly added vehicles are automatically parameterized in the software during the first half-hour of usage.

Read more about how our online vehicle mass measurement works.

With the loaded vehicle mass as input, reliable battery range prediction is possible for electric last mile delivery operations. This allows that during a working day, every single urban utility vehicle will return to a safe charging harbor before its battery is flat.

The following benefits can be anticipated for fleet managers with commercial electric vehicles:

• Reliable, instantaneous vehicle mass at a precision of +/-5%.
• Easy integration into existing telematics solution.
• Automatic setup of new vehicles.
• A battery range estimation that you can trust.
• Increased customer satisfaction.

It may sound counter-intuitive, but 12V battery failures still account for half of roadside assistance requests of electric passenger cars and light commercial vehicles (electric vans and small trucks) according to the 2020 statistics based on 3,4 million assistance requests of ADAC, Europe’s largest motoring association. Due to the impact of COVID19, there have been slightly more battery failures in 2020 as soon as the lockdown ended.

What sounds even more weird is that the share of 12V battery failures is higher (namely 54%) for EVs than for combustion engine based vehicles (46%)! Such differences are due to the fact, that electric vehicles also need a 12V battery for the ability to open the doors, to activate the high voltage battery and to power the onboard computer. All onboard electronics like infotainment, GPS, radio and lights work are powered by this small 12V battery.

Actual project results show that we can reliably predict future events of battery failure through continuous 12/24V battery condition monitoring and allow our customers to change “tired” batteries before they fail – and only those.

Our virtual sensors can also predict fatigue and failure of other electrical and mechanical vehicle components like tires and brakes. We can alert about brake system anomalies and thus reduce risks for your drivers significantly.

Get in touch with us to discuss about your Use case!