SIVALab joint laboratory

SIVALab (Integrity Systems Laboratory for Autonomous Vehicles)

From 2017 to 2025, Renault (Ampere) and Heudi­asyc joint­ly coor­di­nat­ed SIVAL­ab, a lab­o­ra­to­ry spe­cial­ized in local­iza­tion and per­cep­tion sys­tems for autonomous vehicles.

This sci­en­tif­ic and tech­no­log­i­cal part­ner­ship is the result of more than ten years of col­lab­o­ra­tion and builds upon Heudiasyc’s autonomous vehi­cle plat­forms, devel­oped using Renault ZOE vehicles.

Research Areas

The main focus of the SIVAL­ab joint lab­o­ra­to­ry is the integri­ty mon­i­tor­ing of infor­ma­tion derived from per­cep­tion, local­iza­tion, and dig­i­tal maps for Advanced Dri­ver Assis­tance Sys­tems (ADAS) and Autonomous Dri­ving (AD) func­tions in intel­li­gent vehi­cles. The con­cept of integri­ty is based on the com­bi­na­tion and fusion of data from com­ple­men­tary, diverse, and redun­dant sources.

This infor­ma­tion is con­sol­i­dat­ed with­in an envi­ron­men­tal rep­re­sen­ta­tion frame­work known as the World Mod­el, which serves as the inter­face between the per­cep­tion and local­iza­tion mod­ules and the deci­sion-mak­ing module.

The objec­tive of SIVAL­ab is to ensure that the World Mod­el pro­vides reli­able real-time infor­ma­tion for sit­u­a­tion aware­ness and under­stand­ing, enabling deci­sion-mak­ing for ADAS and autonomous dri­ving func­tions. This includes ensur­ing the reli­a­bil­i­ty of per­cep­tion, local­iza­tion, and map­ping infor­ma­tion with­in a spec­i­fied Oper­a­tional Design Domain (ODD).

Projects and collaborations

ESCAPE : Euro­pean project fund­ed by the Euro­pean GNSS Agency (GSA). With­in this project, Heudi­asyc par­tic­i­pat­ed through Renault, one of the project part­ners. The objec­tive was to devel­op a high-integri­ty posi­tion­ing sys­tem by com­bin­ing the new capa­bil­i­ties offered by the Euro­pean Galileo satel­lite nav­i­ga­tion sys­tem with intel­li­gent cam­eras and high-def­i­n­i­tion maps.

TORNADO : French FUI-fund­ed project that inves­ti­gates inter­ac­tions between autonomous vehi­cles and infra­struc­ture to enable mobil­i­ty ser­vices in low-den­si­ty areas. Field tri­als were con­duct­ed in Ram­bouil­let in 2021 under real-world oper­at­ing con­di­tions on pub­lic roads.

ERASMO : project aimed at devel­op­ing and demon­strat­ing an advanced On-Board Unit (OBU) for auto­mat­ed dri­ving. Lever­ag­ing the exper­tise of lead­ing Euro­pean indus­tri­al and aca­d­e­m­ic partners—including Renault Group, Idneo, GMV, Kudan, and Septentrio—the project pro­pos­es a mul­ti-sen­sor solu­tion that com­bines mul­ti-con­stel­la­tion GNSS sig­nals with data from both inter­nal and exter­nal vehi­cle sensors.

Experimentations

• Field tri­als in Ram­bouil­let as part of the TORNADO project.
• Detec­tion and mon­i­tor­ing of dri­ving mode at traf­fic lights.
• Demon­stra­tion of an autonomous vehi­cle on the closed test track in Seville.
• Demon­stra­tions at the IEEE Intel­li­gent Vehi­cles Sym­po­sium (IV 2019).

Key figures

• 3 post­doc­tor­al researchers and 10 engi­neers recruit­ed dur­ing the project period
• 20 Mas­ter’s stu­dents host­ed as interns
• 14 PhD dis­ser­ta­tions com­plet­ed and 5 ongo­ing
• 15 jour­nal pub­li­ca­tions and 55 con­fer­ence papers
• 9 patents and 1 soft­ware pack­age filed
• Around 10 work­shops orga­nized through­out the project period