Altran’s parent, Capgemini, is combining the engineering skills of Altran with its data infrastructure and has launched a service for end to end support for validation and verification of driverless car systems. Altran also includes Cambridge Consultants in the UK which has been developing AI and sensor technologies for autonomous systems.
The validation technology is being used by the maker of Citroen, Peugeot and Opel-Vauxhall cars to manage thousands of petabytes of data from testing the next generation of driverless cars.
“We wanted to work with Capgemini and Altran because of their strong skills in data oriented and cloud- based projects. Participating in a European innovation project for the automotive industry in the field of the connected and autonomous car is very challenging. This collaboration enables us to complete our data collection and processing on schedule, and helps us to deploy innovative solutions for data analysis methods on a hybrid-cloud based solution,” said Jean-Louis Sauvaget, Research & Development Division, Expert car data acquisition and post processing for customer usage in Groupe PSA.
Achieving acceptable levels of safety and reliability to enable autonomous driving will need substantial verification and validation activity. This requires vast numbers of simulations and road tests, generating data at vast scale. Capgemini’s Driving Automation Systems validation service will offer car makers and Tier One suppliers test strategy and orchestration as well as physical and virtual data production and data management. This includes data management services, tool integration and digital continuity, hybrid cloud infrastructure and a full verification and validation as a service.
“Autonomous driving is an exciting outcome of Intelligent Industry. But it is a complex undertaking: achieving acceptable levels of safety and reliability will require substantial verification and validation,” said Franck Greverie, Chief Portfolio Officer at Capgemini and Group Executive Board member. “A single vehicle can generate several hundreds of petabytes of data through driving. But no vehicle can drive enough miles and generate enough data to anticipate how to behave in any scenario on a