Electric car: GPS optimizes battery life by 10%

Researchers at the University of Riverside, California, are developing an algorithm that would allow a GPS to optimize the range of an electric car by 10%, by making it choose the best routes to save its batteries.

In recent years, GPS - whether autonomous or on board - has not only offered a single route when they are given a starting point and a point of arrival. They generally calculate the shortest and fastest route (which are not necessarily the same depending on the type of routes taken) and sometimes also the most economical. That is, the one who will spend the least fuel avoiding cities and preferring lanes limited to 110 km/h to motorways.

But what is practical for a conventional vehicle becomes crucial with an electric car, whose range is still very limited. It is 160 km on average for models currently sold of the Nissan Leaf type, but can vary - usually downwards - depending on elements such as average speed and outside temperature (the colder the longer the range). It is therefore necessary to go further by including these parameters but also others, such as the gradient of the road, the weight of the vehicle, the number of passengers or the traffic jams in real time. These are all characteristics that can vary, from simple to double, the battery life over the same distance and whose calculations must take into account to be sure that there is enough electricity left to go at the end of the journey.

Electric car: a 10% energy gain

This is the subject that researchers from the Centre for Environmental and Technology Research (CE-CERT) at the University of Riverside in California have been studying. They believe that it is possible to increase the range of electric vehicles by 10% by taking these parameters into account and optimizing GPS calculations. The Federal Energy Regulatory Commission (Ferc) of the State of California has just awarded them a grant of $95,000 (equivalent to 77,500 euros) to carry out their work, which goes through the development of a special algorithm and measurements in real conditions.

For this, the four people involved in the project do not start from scratch since they can be based on pre-existing work. "Our team started from the work of Professors Boriboonsomsin and Barth, who developed a skinning system for petrol cars and large diesel trucks," Guoyuan Wu, assistant researcher at CE-CERT and project manager, told Futura-Sciences. "We will develop on this basis a kind of prototype of the new system specially dedicated to electric vehicles. This is particularly useful given their limited autonomy. There is a real way to reduce driver anxiety about this criterion. 

The thrifty GPS for electric cars in prototype

If the CE-CERT team can develop a more complicated and complete algorithm than THAT of GPS, it is because it knows that it can rely on the increasingly powerful computing capacity of processors embedded in the multimedia systems of cars. Some of them even use NVidia's Tegra chips, the same ones that are used in high-end smartphones and touch tablets. This makes it easy for them to assimilate new data and mix traffic information with a topographic map, weather forecasts and elements directly from the vehicle, such as the power of the air conditioning or the number of passengers on board.

Battery consumption measurements during traffic tests will be compiled into a model on which this new algorithm will be based. It will serve as the basis for a prototype of a new-generation touchscreen navigation system that can be installed natively in an electric vehicle. "Our role is to develop this technology, and then it will be manufacturers or even consumers who will have the power to decide whether such a system can be commercialized. At the moment, our relationship with manufacturers boils down to electric car loans for our measurements," says Guoyuan Wu. It is therefore too early to consider a concrete commercialization of this intelligent navigation system.

Especially since the prototype developed is only valid for California. However, Mr. Wu is reassuring: "We work with an American data provider but it would be very easy to transpose this model to other continents or countries like Europe and France, where there are also such suppliers."

This could greatly decide those interested in buying an electric vehicle and who do not wish to suffer the anguish of the breakdown between two charging stations.

Post a comment