ELECTRIC VEHICLES Algorithm EV specialist


cooperates with oil producer to trial AI based prediction


technology to reduce range anxiety.


R


❱ ❱ Tests are showing that artificial intelligence algorithms could improve range prediction accuracy to single figure kilometers


ange anxiety, the doubt suffered by electric vehicle (EV) owners that they have enough charge to reach their destination, is a strong influence in making the decision to switch from the internal combustion engine or hybrid power. There is some indication of remaining charge and how that


translates to distance but it’s based on pre-configured assumptions and doesn’t alter with the style of driving. Now, EV specialist Spark EV Technology is trialling its advanced range


prediction system, designed to overcome range anxiety and give confidence in prediction accuracy. The company joined forces with BP to perform the trials, using 5 of its employees to drive 10,000km each using the Artificial Intelligence algorithms supplied by Spark.


TESTING AI PREDICTIONS During the test, the BP drivers entered their proposed journey into Spark’s smartphone application to record data and make comparisons with the onboard range display. They then received personalised advice on whether they could complete it – based on live data, driving style, urban/country routes, previous trips and charge point locations. The algorithms learn with every journey, continually improving the accuracy of personalised journey predictions. When it comes to EVs, incorrect range predictions impact drivers in two ways.


Underprediction, where the actual journey uses more energy than forecast, leads to range anxiety, with drivers worried about being able to complete their journey. Its opposite, overprediction, when a journey actually uses less range than projected, contributes to time anxiety, as drivers unnecessarily charge their EV or put off trips altogether. EVs today are poor at estimating how far their battery will take them, and in real world tests can significantly underestimate or overestimate the energy needed for a journey. Urban driving can be 2 or 3 times as efficient as motorway driving, and yet the vehicle uses the most recent driving efficiency as a guide to the remaining range. This means vehicles can significantly over or underestimate range when cars change from urban to motorway driving, or vice-versa. During Spark’s trial with BP it was observed that one car could have driven 45 miles more than its displayed range, and on another journey an extra 21 miles more range had to be used to complete the journey than the car predicted. In comparison, Spark intelligent range prediction uses data on where the car is going, not where it has been, and for the same journeys predicted the range to an accuracy of just 3.8 miles and 0.4 miles respectively.


20 /// Automotive Test & Validation Vol 2 No. 2 Anxiety Range


Automotive Test & Validation Vol 2 No. 2 automotiveTV@concordepublishing.com


Editorial


Direct Line +44 (0) 20 7863 3078 Editor Jonathan Newell BSc


jon.newell@concordepublishing.com Advertising


Direct Line +44 (0) 20 7863 3077


Advertising Manager John Harvey john.harvey@concordepublishing.com


Managing Director Paul Williams


paul.williams@concordepublishing.com Concorde Publishing Ltd


100 Borough High Street, London SE1 1LB, UK Main office +44 (0) 20 7863 3079 info@concordepublishing.com


www.environmentalengineering.org.uk


Automotive Test & Validation is produced by Concorde Publishing Ltd


Coming up in this series... & Testing+ & Test Houses


ConcordePublishingLtd Publications by engineers for professional engineers


© Concorde Publishing Ltd 2019 This publication is copyright under the Berne Convention and the international Copyright


Convention. Apart from any fair dealing for the purpose of private study, research, criticism or review, as permitted under the Copyright, Designs and Patents Act, 1988, no part may be reproduced, stored in a retrieval system, or transmitted in any form or by any means,


electronic, electrical, chemical mechanical, photocopying, recording or otherwise, without the prior permission of the copyright owners. Unlicensed multiple copying of the contents of this publication is illegal. Inquiries should be addressed to: The Publisher, Concorde Publishing Ltd, 100 Borough High Street, London SE1 1LB, UK.


Views expressed are those of the authors and are not necessarily endorsed by


Concorde Publishing Ltd. Data and conclusions developed are for information only and are not intended for use without independent substantiating investigation by the potential user.


EXTRA


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24