The annu­al UTC AI meet­ing took place on Fri­day Sep­tem­ber 26, 2025 at the Daniel Thomas Inno­va­tion Cen­tre. This third edi­tion focused on arti­fi­cial intel­li­gence and the indus­try of the future. Speak­ers includ­ed Renault Group, Psy­cle, Alfi and Feal­inx address­ing the sub­ject of AI for pro­duc­tion. The day was also ded­i­cat­ed to SME companies.

On the agen­da for this 3rd UTC UTC-AI meet­ing, sup­port­ed by the UTC Foun­da­tion for Inno­va­tion, par­tic­i­pants learned a great deal about AI in indus­try, the pos­si­bil­i­ties between pre­dic­tive AI and gen­er­a­tive AI, as well as AI for pro­duc­tion and AI from the per­spec­tive of large groups and SMEs. Emmanuelle Abis­set-Cha­vanne, pro­fes­sor at Arts et Métiers Paris­Tech-ENSAM, and Alexan­dre Durupt, senior lec­tur­er at UTC-Com­pieg­ne, pre­sent­ed an inter­ac­tive quiz sum­maris­ing the main sur­veys and stud­ies con­duct­ed on the deploy­ment and use of AI by SMEs in France. “Our aim here is to demys­ti­fy AI and gain a bet­ter under­stand­ing of how to deploy it and how it can ben­e­fit indus­try and our pro­fes­sions. AI is cer­tain­ly an atyp­i­cal col­lab­o­ra­tor that has fall­en vic­tim to its own hype. AI is not mag­ic. Com­pa­nies must first ask them­selves whether their need is real, whether they have the data to engage with AI, and to do so, they must also know how to engage in dig­i­tal trans­for­ma­tion and devel­op a cul­ture of use. We must learn to use AI, to work with it, and to train our­selves in its use in a respon­si­ble man­ner,‘ insist­ed the first two speak­ers before an audi­ence of atten­tive engi­neers and entre­pre­neurs. “Struc­tur­ing data, labelling it, and above all, hav­ing con­fi­dence in it, are the first steps in dig­i­tal trans­for­ma­tion, which is at the heart of the sen­si­ble use of AI by com­pa­nies.” These com­ments reflect the UTC Rober­val lab­o­ra­to­ry’s ongo­ing com­mit­ment to bring togeth­er aca­d­e­m­ic research and indus­tri­al needs in order to build the future togeth­er. Par­tic­i­pants also had the oppor­tu­ni­ty to vis­it the lab­o­ra­to­ry’s AMS (Advanced Man­u­fac­tur­ing Sys­tem) and Qua­tri­um (CETIM) platforms.

Proof by example at Renault

Meriem Lafou, AI archi­tect and Antoine Leblanc, AI and Data expert at Renault Group, also rep­re­sent­ed the SIA (Société des ingénieurs de l’au­to­mo­bile). At the Renault Group, AI is now estab­lish­ing itself as a valu­able tool for deci­sion-mak­ing and oper­a­tions. It rep­re­sents a real lever for indus­tri­al per­for­mance. “It is inte­grat­ed into con­crete tools such as Mae­stro, which uses gen­er­a­tive AI to sup­port main­te­nance oper­a­tors. The result: a reduc­tion of more than 40% in the time spent on machine break­downs. AI is not an end in itself; our pri­or­i­ty is to meet con­crete indus­tri­al needs with explain­able and deter­min­is­tic solu­tions. That is why we favour sym­bol­ic AI. How­ev­er, when this does not pro­vide a sat­is­fac­to­ry answer, we use con­nec­tion­ist AI, ensur­ing that it remains con­trolled and jus­ti­fied by its use,‘ explains Meriem Lafou. An intel­li­gent mon­i­tor­ing solu­tion has been imple­ment­ed on stamp­ing press­es, based on accelerom­e­ters capa­ble of trans­mit­ting data in real time. “As soon as a crit­i­cal thresh­old is crossed, an alert is auto­mat­i­cal­ly gen­er­at­ed,” explains Meriem Lafou. “This allows us to antic­i­pate break­downs, pre­pare inter­ven­tions and quick­ly detect any poten­tial anom­alies.” AI has also been used to pre­dict gas and ener­gy con­sump­tion in Renault’s car assem­bly fac­to­ries, enabling them to achieve a 20% increase in pro­duc­tiv­i­ty. The round table then turned to the ques­tion of the appli­ca­tions and ben­e­fits of AI in indus­try, with Yves Grand­valet, Direc­tor of Research at the CNRS, Meriem Lafou from the Renault Group, Philippe Boutin­aud, Tech­ni­cal and Inno­va­tion Direc­tor at Feal­inx, and Nathalie Greenan, Pro­fes­sor at the CNAM. The theme of the next edi­tion will be ‘AI and process­es’ in March 2026.

Accompanying corporate transformation projects

Muriel Wind­holtz is in charge of oper­a­tions for the Qua­tri­um accel­er­a­tion net­work ded­i­cat­ed to local busi­ness trans­for­ma­tion projects, planned to start on Novem­ber 14, 2025. She is also respon­si­ble for busi­ness rela­tions at UTC. This project was launched by CETIM (Sen­lis) in 2023 and joined by UTC-Com­pieg­ne in ear­ly 2025 thanks to fund­ing from the Region and Europe. The project aims to pro­vide busi­ness­es with sup­port from spe­cial­ists in order to estab­lish a per­son­alised diag­no­sis, define an action plan and assist in the choice of solu­tions through to indus­tri­al­i­sa­tion. “The Qua­tri­um project sup­ports indus­tri­al SMEs in their dig­i­tal and eco­log­i­cal trans­for­ma­tion. Demon­stra­tion plat­forms for exist­ing tech­nolo­gies are avail­able to iden­ti­fy solu­tions that best meet the com­pa­ny’s needs,” she says. What is UTC’s spe­cif­ic role? “For our part, we pro­vide solu­tions that com­ple­ment those of oth­er part­ners. This year, we have set up a ded­i­cat­ed space at UTC to present new tech­nolo­gies to local man­u­fac­tur­ers. While CETIM spe­cialis­es in the mechan­i­cal indus­try, at UTC we tar­get a broad­er spec­trum of indus­tries such as chem­i­cals and agribusi­ness. The idea revolves round devel­op­ing con­nect­ed tools in man­u­fac­tur­ing process­es, such as a smart con­vey­or line,” con­cludes Muriel Windholtz.

KD