Recherchez sur le site

Generic filters
Exact matches only
  • Heuristics and Diagnostics for Complex Systems (Heudiasyc) – UMR CNRS 7253

    Foun­ded in 1981 and asso­cia­ted with the CNRS from the start, the mis­sion of the UTC-Heu­dia­syc research unit is to car­ry out research into auto­ma­tion, deci­sion-making, ima­ging and com­pu­ting, taking human fac­tors into account.


    The scien­ti­fic pro­ject deve­lo­ped within Heu­dia­syc is based on the syner­gy bet­ween ups­tream, basic research and fina­li­zed, applied research, to meet society's main chal­lenges: safe­ty, trans­por­ta­tion, STIC (Science and Tech­no­lo­gy for Infor­ma­tion and Com­mu­ni­ca­tion), envi­ron­ment and health in close col­la­bo­ra­tion with various pro­fes­sio­nal part­ners, nota­bly indus­tria­lists. Seve­ral plat­forms and demons­tra­tors have been assem­bled in the labo­ra­to­ry and they illus­trate per­fect­ly the scien­tists' desire to confront their basic research work with com­plex appli­ca­tion areas.

    The objec­tive is to pro­vide the means of repre­sen­ting, ana­ly­sing and control­ling tech­ni­cal sys­tems sub­ject to cri­te­ria and constraints expres­sed in scien­ti­fic, tech­no­lo­gi­cal, eco­no­mic terms and in terms of impact on socie­ty and the human population.

    Teams and research topics

    Heu­dia­syc is deve­lo­ping its acti­vi­ties with 3 research teams:

    • Know­ledge, uncer­tain­ty, data (CID),
    • Safe­ty, com­mu­ni­ca­tion, opti­mi­za­tion (SCOP),
    • Robo­tic sys­tems in inter­ac­tion (SyRI).


    Patents and licences (2011−2016)

    • 10 patents filled and issued, invol­ving mem­bers of the laboratory
    • 4 inter­na­tio­nal exten­sions, 3 exten­sion requests in progress
    • 5 soft­wares depo­si­ted with the APP (Agence de pro­tec­tion des programmes)


    Imple­men­ta­tion of the methods deve­lo­ped takes place on expe­ri­men­tal test rigs in the labo­ra­to­ry: ins­tru­ment fit­ted vehicles, drone fleets and mobile robots, sen­sor arrays, vir­tual rea­li­ty (VR), rail­road sur­veillance, inter-vehicle com­mu­ni­ca­tion, tac­tile tables, dis­tri­bu­ted com­mu­ni­ca­tion, etc.


    At local level

    Heu­dia­syc acti­ve­ly par­ti­ci­pates at all levels of uni­ver­si­ty life (trai­ning, research, pro­mo­tion, etc.). Col­la­bo­ra­tion with the various UTC labo­ra­to­ries takes place main­ly within the fra­me­work of two trans­ver­sal structures:

    • Labex MS2T, "labo­ra­toire d'excellence Maî­trise des sys­tèmes de sys­tèmes tech­no­lo­giques", of which Heu­dia­syc is the lea­der, in asso­cia­tion with two research units of UTC and CNRS (BMBI, Rober­val), since 2011.
    • FR CNRS SHIC, research fede­ra­tion "Sys­tèmes hété­ro­gènes en inter­ac­tion", since 2009.

    At regional level

    The Hauts-de-France Region (Picar­dy Region before 2016) pro­vides research sup­port in the form of fun­ding for three-year pro­jects. These pro­jects are gene­ral­ly co-finan­ced with the FEDER (Euro­pean Eco­no­mic and Regio­nal Deve­lop­ment Fund).

    • 26 regio­nal pro­jects bet­ween 2011 et 2016

    On a national level

    The labo­ra­to­ry par­ti­ci­pates in the acti­vi­ties and ani­ma­tion of seve­ral CNRS research groups: MACS, Robo­tics, Ope­ra­tio­nal Research, IGRV, ISIS, RSD, PréG­dR IA.

    Heu­dia­syc is also invol­ved in many orga­ni­za­tions or natio­nal autho­ri­ties: CNRS (scien­ti­fic advi­sor at INS2I), ANR (mem­ber of com­mit­tees pro­gram eva­lua­tion), natio­nal autho­ri­ties (CNU 61 and 27).

    • 25 pro­jects fun­ded by the ANR bet­ween 2011 and 2016.

    At an international level

    The labo­ra­to­ry main­tains close rela­tions with many Euro­pean and inter­na­tio­nal part­ners, through pro­jects, exchanges of resear­chers, joint the­sis super­vi­sion (Alge­ria, Ger­ma­ny, Aus­tra­lia, Cana­da, Chi­na, Uni­ted States, Ita­ly, Leba­non, Poland, Thai­land, etc. .).

    Among the many col­la­bo­ra­tions, we will par­ti­cu­lar­ly mention:

    • The French-Mexi­can labo­ra­to­ry for com­pu­ter science and auto­ma­tic control sys­tems (LAFMIA), the first Fran­co-Mexi­can inter­na­tio­nal joint unit (UMI) of the CNRS, in part­ner­ship with the GIP­SA-lab labo­ra­to­ry and the Mexi­can research cen­ter CINVESTAV.
    • Joint pro­ject with the Fran­co-Chi­nese labo­ra­to­ry of com­pu­ter science, auto­ma­tion and applied mathe­ma­tics (LIAMA) and "Key Labo­ra­to­ry of Machine Per­cep­tion" of Peking Uni­ver­si­ty (Chi­na).
    • 30 forei­gn visi­ting pro­fes­sors wel­co­med bet­ween 2011 and 2016 (indi­vi­dual stay> 1 month).
    • 25 inter­na­tio­nal pro­jects and 24 Euro­pean pro­jects bet­ween 2011 and 2016.

    Industrial partners

    SIVA­Lab (Sys­tèmes intègres pour le véhi­cule auto­nome), crea­ted in 2017 by Renault and Heu­dia­syc, is a joint labo­ra­to­ry spe­cia­li­zing in loca­li­za­tion and per­cep­tion sys­tems for auto­no­mous vehicles. This scien­ti­fic and tech­no­lo­gi­cal part­ner­ship is laun­ched for four rene­wable years. It is the result of a part­ner­ship of more than ten years and is based on Heudiasyc's auto­no­mous vehicle plat­forms, deve­lo­ped on the basis of the Renault Zoé. 

    • 164 theses fun­ded bet­ween 2011 et 2016, inclu­ding 25 with companies.


    Within the context of regio­nal pro­jects, Heu­dia­syc is par­ti­ci­pa­ting in 2 spe­cia­li­za­tions for the Picar­die Region's stra­te­gic plan for research :

    • mobi­li­ty fac­tors and urban life,
    • smart vehicles and a sys­te­mic ana­ly­sis of pas­sen­ger and freight mobility.

    And two trans­verse priorities:

    • health, life-styles
    • edu­ca­tion, lear­ning, technologies

    At the natio­nal level, the labo­ra­to­ry has accep­ted the chal­lenge of the govern­ment incen­tive "Invest­ments for the Future"; it both:

    • holds the "Control of Tech­no­lo­gi­cal Sys­tems-of-Sys­tems" Labo­ra­to­ry of Excel­lence award (Labex MS2T),
    • and co-ordi­nates the mobile robo­tics sec­tion of the "Robo­tex" equip­ment of excel­lence.

    UTC-Heu­dia­syc is an active par­ti­ci­pant in the two glo­bal com­pe­ti­tive clusters:

    • "i‑Trans", which covers inno­va­tive land trans­por­ta­tion sys­tems (Nord-Pas-de-Calais, Picardie),
    • Sys­te­ma­tic Paris-Région, clus­ter dea­ling with soft­ware and com­plex sys­tems (Ile-de-France Region, Grea­ter Paris).

    It is invol­ved in seve­ral Research Net­work Groups and col­la­bo­rates with a num­ber of labo­ra­to­ries both in France and abroad, in nume­rous research projects.

    ANR projects

    RéCIF an acro­nym for belief net­works desi­gned to assess Ope­ra­ting Safe­ty para­me­ters for rail-road infra­struc­tures quan­ti­ta­ti­ve­ly. The aim is to inves­ti­gate the facts and to pro­pose solu­tions to model rail-road infra­struc­tures using socal­led belief networks.

    DIADEM, an acro­nym for Dyna­mic DIAg­nos­tics and Pro­vi­sio­nal Main­te­nance for Train moun­ted on-board sys­tems. The aim of this pro­ject is to deve­lop diag­no­sis and pro­gnos­tics tools of the three 'sen­si­tive' train devices (air-condi­tio­ning, bra­ke­sys­tems and door ope­ra­tions) that would enable real time poten­tial fore­cas­ting before fai­lure. This per se adds an impro­ve­ment to main­te­nance work and to time-table compliance.

    Eco­pack is a pro­ject that focuses on new forms of col­la­bo­ra­tive work that have become pos­sible by a com­pu­ter-inten­sive socio-tech­ni­cal envi­ron­ment inclu­ding various devices (tables, tac­tile pads, smart­phones, PCs). One of the objec­tives is to offer a sha­red appli­ca­tion acces­sible for each col­la­bo­ra­tor via these devices.

    VEGAS is a pro­ject with the objec­tive of desi­gning, deve­lo­ping and asses­sing a vir­tual labo­ra­to­ry based on co-simu­la­tion, with two types of simu­la­tor. The first is an ERTMS simu­la­tor that models the system's func­tions. The second simu­la­tor is a tele­comm sys­tem simu­la­tor that allows model­ling of abso­lu­te­ly any form of tele­com­mu­ni­ca­tions, from the phy­si­cal layer to the appli­ca­tion layer.

    FUI projects

    SERA : as we observe the deve­lop­ment of aids to dri­ving, there are new oppor­tu­ni­ties for an imple­men­ta­tion of aug­men­ted rea­li­ty tech­niques (com­pu­ter gene­ra­ted images super­im­po­sed on rea­li­ty) in a vehicle for the pur­pose of impro­ving ove­rall safe­ty fac­tors and drive-pas­sen­ger com­fort. The pro­ject calls for the assem­bly of a pro­to­type vehicle inte­gra­ting a came­ra image acqui­re­ment sys­tem and a wind-shield opti­cal pro­jec­tion sys­tem to add the com­pu­ter gene­ra­ted images (aug­men­ted rea­li­ty) rela­ted to dri­ver-use­ful data – the pro­to­type pre­fi­gures self-dri­ve­ve­hicle 'cock­pits'.

    European projects

    CoMo­SeF (Co-ope­ra­tive Mobi­li­ty Ser­vice for the Future) aims at crea­ting ser­vices and equip­ment nee­ded for a large imple­men­ta­tion of co-ope­ra­tive appli­ca­tions in trans­port, com­plying with the ITS Action Plan as well as with natio­nal stra­te­gies of the other Euro­pean partners.

    MASH (MAs­sive Sets of Heu­ris­tics) is a pro­ject for the desi­gn of com­pu­ter sys­tems that assist and enhance col­la­bo­ra­tive deve­lop­ment of new auto­ma­ted lear­ning sys­tems to ensure suc­cess­ful com­plex task management.

    VERVE is a joint inter­na­tio­nal research ven­ture, invol­ving UTC-Heu­dia­syc and the Tech­ni­cal Uni­ver­si­ty (TU) Ilme­nau at Thü­rin­gen, Ger­ma­ny. The scien­ti­fic objec­tive assi­gned to the pro­ject is the deve­lop­ment of a new concept for a "smart" all elec­tric vehicle (SEV) that will contri­bute to smart, eco­lo­gi­cal­ly safe, trans­port sys­tems to be tes­ted in both the Ger­man Thü­rin­gen and French Picar­die regions.

    Picardie regional projects

    CIME (Contex­tual Inter­ac­tions for Mobi­li­ty in Edu­ca­tion) is a research pro­gramme that aims a crea­ting com­pu­ter based lear­ning envi­ron­ments that are context sen­si­tive to be imple­men­ted in mobile devices such as e‑pads or smart­phones to be used to assist visits to museums and other heri­tage site.

    SEDVAC (on-board vehicle dyna­mics assess­ment and dri­ver aid) is a pro­gramme desi­gned to deve­lop a sys­tem that will inte­grate vehicle envi­ron­ment data (for example, maps, tyre-road sur­face forces …) so as to eva­luate the dyna­mic beha­viour of the vehicle and its impact on dri­ver beha­viour …and to pos­si­bly alert the dri­ver should a risk-pro­ne/ab­nor­mal situa­tion occur.

    ECHOPEDIA is a pro­gramme that aims at deve­lo­ping methods to res­tore and seg­ment 3D pedia­tric echo-car­diac images for the pur­pose of explo­ring pos­sible conge­ni­tal cardio-pathologies.

    ARAKIS is a pro­gramme to deve­lop simu­la­tions of various hazar­dous work condi­tions (dan­ge­rous sub-contrac­tor work in high risk zones (in 'Seve­so' cate­go­ry sites notably).

    Zoom on 2 projects

    The concept of wire­less sen­sor net­works (WISNs) leads to new and effi­cient means to limit the nega­tive impact of agri­cul­tu­ral prac­tice on the envi­ron­ment. Pro­fes­sio­nals in the agro-sec­tor – if they choose to imple­ment Agro­Sens – will be in a posi­tion to opti­mise use of fer­ti­li­sers, etc., and thus help to pre­serve the natu­ral resources of the Earth. As Chal­lal YACINE, senior lec­tu­rer at the Uni­ver­si­ty UTC puts it, "Wire­less sen­sor net­works ensure data retrie­val and trans­mis­sion in a reliable and secure man­ner to the system's super­vi­so­ry and control sta­tion".

    Data will feed into a deci­sion aid sys­tem for agri­cul­ture (or for envi­ron­men­tal moni­to­ring). Abdel­mad­jid BOUABDALLAH, the UTC pro­fes­sor in charge of this pro­ject, adds that "we were faced with a real chal­lenge inter­ms of ener­gy saving and secure rou­ting of the data coming from sen­sors pla­ced in often high­ly varied soil cha­rac­te­ris­tics and envi­ron­ments."

    A WISN infra­struc­ture allows ope­ra­tors to bene­fit from conti­nuous super­vi­sion of the state of the milieu (air tem­pe­ra­ture, hygro­me­tric level, soil humi­di­ty, solar radia­tion), the state of the crops (leaf tem­pe­ra­ture), the­re­by gai­ning in opti­mi­sa­tion of use of resources (water, fer­ti­li­sers and pes­ti­cides), like­wise hel­ping attain bet­ter mana­ge­ment control of irri­ga­tion, and impro­ving crop yield in terms of quan­ti­ty and qua­li­ty and, last but not least, hel­ping to pre­dict pos­sible plant diseases.

    If you can gain half a second when you rea­lize there's an obs­tacle ahead of you on a road, then you have seve­ral metres extra dis­tance to brake and in the end this may save lives (inclu­ding yours). The research team hea­ded by Phi­lippe Bon­ni­fait, cal­led "Auto­ma­tion, Onboard Sys­tems, Robo­tics" with the Heu­dia­syc Labo­ra­to­ry, aims at impro­ving vehicle mobi­li­ty and safe­ty by intro­du­cing so-cal­led smart cars.

    "The ques­tion", reflects Phi­lippe "is how can a dri­ver become a super­vi­sor?" This does not mean remo­ving human beings from the dri­ving seat, but rather to make the most of the machine's built-in capa­ci­ty to control and regu­late motion and navi­ga­tion, in order to free the driver's mind for deci­sion-making pro­cesses, as and when needed.

    "Dri­ving can be sum­ma­ri­sed as 2 main ope­ra­tions: per­cep­tion of theen­vi­ron­ment in which the vehicle is moving and then orde­ring the vehicle to move, appro­pria­te­ly, as desi­red. So, if you want to make the vehicle more intel­li­gent, i.e., smar­ter, you must give it the capa­ci­ty to per­ceive the out­side world"; Phi­lippe Bonnifait's teams have ins­tal­led on-board sen­sors : hyper or ultra­so­nic fre­quen­cy radars (used for example as par­king aids), optic tele­me­try (lasers, came­ras), loca­li­sa­tion sys­tems (GPS). The chal­lenge the­reaf­ter is to pro­cess the data in real time. "At UTC, we are wor­king on how to bring the data toge­ther and create phy­si­cal models as an appli­ca­tion of the gene­ral, so-cal­led, sepa­ra­tion prin­ciple. In the first ins­tance, we deal with ques­tions that relate to per­cep­tion, fol­lo­wed by those that relate to control func­tions."

    Howe­ver, mea­su­ring the out­side world is not so easy as we can ima­gine: "first­ly, you have to wor­kout the degree of relia­bi­li­ty of the data pro­vi­ded by the sen­sors, to avoid taking a deci­sion that in fact is using false infor­ma­tion". The constraints are part­ly phy­si­cal – sen­sors have a limi­ted range of inves­ti­ga­tion; came­ras work less well (or not at all) at night; lasers are dis­tur­bed by rain dro­plets. The sum­med effect of these uncer­tain­ties cor­res­ponds to what scien­tists call ambient or back­ground "noise", signals that deform data – and this is an area of research that is of par­ti­cu­lar inter­est to the UTC scien­tists. "We are cur­rent­ly wor­king on schemes that allow us to mode­lise the uncer­tain­ties; from these we esta­blish a degree of confi­dence that can be assi­gned to each result."

    Smart cars are not just for eve­ry­day, nor­mal dri­vers. They could be exten­ded to han­di­cap­ped per­sons or to those who are get­ting too old to drive. We can ima­gine people dri­ving in per­fect­ly safe condi­tions even though they can not turn their heads left or right or can not see too well. They too should be able to bene­fit from the onboard smart concept, shouldn't they?

    Contact and documentation

    Contacts de la recherche à l'UTC

    Direc­teur du labo­ra­toire Heu­dia­syc
    Phi­lippe Bon­ni­fait
     +33 (0)3 44 23 44 81
    À lire dans Interactions

    « Une vie pro­fes­sion­nelle, ça se construit pas à pas »

    Agir sur le monde en chan­geant le regard sur l’innovation

    Desi­gn sonore made in UTC sur Radio France

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