Simon Col­lot, AI (arti­fi­cial intel­li­gence) expert at Sopra Ste­ria, deliv­ered a lec­ture in Sep­tem­ber at UTC to demys­ti­fy the sub­ject, invit­ing the audi­ence to join him on a tour of the oppor­tu­ni­ties opened up by new AI-based inno­va­tions. Here is sum­ma­ry of his analy­sis and update.

 

Cur­rent­ly used in the ser­vice sec­tor and in var­i­ous indus­tri­al sec­tors, arti­fi­cial intel­li­gence has become increas­ing­ly present in a day-to-day pro­fes­sion­al and pri­vate spheres. The obsta­cles that remain are more human than tech­no­log­i­cal. « The chal­lenge lies every bit as much in writ­ing the pro­gram­ming bricks as in con­vinc­ing the col­leagues in an enter­prise that they can count on gains in effi­cien­cy and also to reas­sure them as to the result­ing divi­sion of labour between men and machines: one of the roles of Sopra Ste­ria being to accom­pa­ny peo­ple in this accul­tur­a­tion”, explains Simon Col­lot who has notably been assigned advi­so­ry mis­sions for and with the ener­gy sec­tor engi­neers and deci­sion-mak­ers. Whilst these AI derived inno­va­tions are often seen as a threat, expert Col­lot is con­vinced, he says, that Machines will assist Men but not replace them. “Automat­ing cer­tain process­es will enable work­ers and oper­a­tives to con­cen­trate on high­er added val­ue tasks and to be assist­ed in doing so”. Increas­ing­ly “agile” solu­tions With chat­bots, spe­cif­ic trade-ori­ent­ed solu­tions, auto-diag­no­sis in the aero­space sec­tor, we see AI is unavoid­able when it comes to propos­ing new ser­vices to clients and to gain in pro­duc­tiv­i­ty in an increas­ing­ly com­pet­i­tive com­mer­cial con­text. It is a new tech­nol­o­gy that relies essen­tial­ly on imple­ment­ing three key stages: a sit­u­a­tion stats and descrip­tion thanks to analy­sis of com­plex data, pre­dict­ing the next events, using mod­els and – for the more advanced solu­tions – solv­ing and cor­rect­ing the prob­lems as and when detect­ed. “With AI the solu­tions are nev­er ‘frozen’ and with auto-learn­ing sys­tems they can be improved and adapt­ed to fit a chang­ing envi­ron­ment”, adds our engi­neer, spe­cial­ist in com­put­er sci­ences and appli­ca­tions. At this point in time, com­put­ers can now imi­tate human con­ver­sa­tion very close­ly. This poten­tial­i­ty can be of high inter­est to cus­tomer ser­vices – cer­tain solu­tions have already been test­ed and oth­ers will come in the near future. After on-line mes­sage sys­tems han­dled by automats, you will soon be able to call up an automat. Banks are among the most advanced enter­pris­es using AI. “One of our cus­tomers has pro­posed extend­ing the weekly/monthly cash with­draw­al lim­it out­side the account holder’s open­ing hours, or to engage in a dia­logue with a robot and open a new account in a total­ly auto­mat­ed man­ner”. Although the elec­tric pow­er com­pa­nies have been lat­er starters in this fields, they are also engaged in the AI rev­o­lu­tion. Their cus­tomer advi­so­ry ser­vices will soon able to rely on a tech­ni­cal assis­tance to pro­pose per­son­alised con­tracts to each consumer/customer. These new ser­vices will allow the enter­pris­es to have a rapid ROI (return on invest­ment) but oth­er long term evo­lu­tions are cur­rent­ly being stud­ied in our lab­o­ra­to­ries. Decen­tral­ized, renew­able ener­gy pro­duc­tion in tomorrow’s ‘smart cities’ will be includ­ed in these major work pro­grammes. Design of ‘smart’ net­works is, indeed, very nec­es­sary to be able to bal­ance sup­ply and demand. News from AI-inten­sive sec­tors will be dense … Sopra Ste­ria has the inten­tion to sat­is­fac­to­ri­ly accom­pa­ny its cus­tomers as these evo­lu­tion­ary trends come to be.