Energy-efficient artificial intelligence for electric mobility

On Octo­ber 2, 2026, the fifth edi­tion of the Arti­fi­cial Intel­li­gence Forum was ded­i­cat­ed to the theme ‘AI for elec­tric mobil­i­ty’, with a par­tic­u­lar focus on energy-efficiency.

This event aimed at fos­ter­ing a con­struc­tive dia­logue between the worlds of research and indus­try in order to explore new avenues for inno­va­tion is organ­ised with the sup­port of the UTC Foun­da­tion and the UTYC-SAFE AI Chair, in part­ner­ship with GIS S.mart, which brings togeth­er 35 high­er edu­ca­tion and research insti­tu­tions spread across main­land France.

S.mart sig­ni­fie Sys­tems Man­u­fac­tur­ing Aca­d­e­mics Resources Tech­nolo­gies. Cf. S.mart – Sys­tem Man­u­fac­tur­ing Aca­d­e­mics Resources Technologies

Why this theme? “Ini­tial­ly, we planned to ded­i­cate this edi­tion to the top­ic of ‘bat­ter­ies and AI’, but very quick­ly, togeth­er with my colleagues—including Ben­jamin Quost and Sébastien Destercke—we decid­ed to broad­en it to include elec­tro­mo­bil­i­ty, on order to cov­er the entire elec­tric pow­er­train and engage a wider aca­d­e­m­ic and indus­tri­al net­work”, explains Nico­las Damay, one of the co-organ­is­ers and a lec­tur­er-researcher at UTC’s Rober­val Laboratory.

Who were some of the speak­ers? “Notably, we heard Emmanuel Batista from Alstom, a part­ner com­pa­ny with whom we are work­ing on bat­ter­ies, with the aim of ensur­ing max­i­mum dura­bil­i­ty. Ale­jan­dro Fran­co, for his part, advo­cates the use of AI, par­tic­u­lar­ly to sim­u­late com­plex phys­i­cal mod­els (based on val­i­da­tion data) in order to speed up sim­u­la­tions. Final­ly, Ben­jamin Quost is work­ing with Renault on the ener­gy opti­mi­sa­tion of elec­tric vehi­cles by com­bin­ing phys­i­cal knowl­edge (bat­tery, invert­ers, motors) with com­plex envi­ron­men­tal data,” he emphasises.

What avenues have been out­lined? “All the speak­ers empha­sised the need for a mea­sured approach. Emmanuel Batista sug­gest­ed using AI to process large vol­umes of data from train fleets that are con­tin­u­ous­ly mon­i­tored. He demon­strat­ed that, if the prob­lem is prop­er­ly framed, a rel­a­tive­ly sim­ple neur­al net­work is suf­fi­cient to iden­ti­fy indi­ca­tors of age­ing, such as a battery’s elec­tri­cal resis­tance. Physics First is the watch­word of ’Ale­jan­dro Fran­co. In his view, AI must be ground­ed in human and phys­i­cal knowl­edge to be effec­tive,” he explains.

And every­one empha­sised the role of humans, par­tic­u­lar­ly dur­ing the round-table dis­cus­sion. “Con­trary to alarmist talk of autonomous ‘super­in­tel­li­gence’, humans remain, in the eyes of all the speak­ers, essen­tial for prop­er­ly defin­ing the prob­lem, guid­ing AI and seek­ing solu­tions. And when it comes to edu­ca­tion, they all empha­sise the impor­tance of main­tain­ing a sol­id foun­da­tion and a crit­i­cal mind­set amongst young peo­ple and oper­ate on the prin­ci­ple that any solu­tion must be explain­able, sub­stan­ti­at­ed and jus­ti­fied”, con­cludes Nico­las Damay.

A suc­cess­ful event, giv­en the high­ly spe­cialised nature of the top­ic, with dozens of peo­ple attend­ing in per­son – includ­ing PhD stu­dents – and over 1 000 views on the YouTube platform.

Watch the lec­tures and the round-table dis­cus­sion at: www.youtube.com/live/Ymfcg5l2ews

MSD

Le magazine

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram