SpirOps AI

SpirOps is an independent Artificial Intelligence lab since 2004. It's a unique model that combines the freedom of research subjects (whether applied or fundamental) with industrial development for application purposes with partners such as Stellantis, SNCF, IRT Railenium, Alstom, Bosch, Thales, RATP...

CROWD

Since 2007, SpirOps has been developing a crowd simulator in which each agent has his or her own desires, personal characteristics and subjective knowledge of the world.

The agents evolve in automatically-generated cities, on public transport, in tourist attractions or in offices.

SELF-DRIVING CAR

Since 2015, SpirOps has been working for Stellantis on the development of Level 3 and 4 autonomous car prototypes.

This includes data acquisition and fusion, intelligent and explainable decision-making, simulation of complex behaviors... to offer the safest and most comfortable driving possible.

PROCEDURAL ANIMATION

Since 2009, SpirOps has been developing a procedural animation engine in which movements are generated in real time, according to a set of physical rules.

This creates realistic animations that can automatically adapt to a wide variety of humanoid skeletons (video games, VR, etc.).

MEMORY & DIALOGUE

Part of SpirOps R&D is dedicated to the development of an artificial memory. This can be incorporated into a simulated agent, enabling it to acquire information in real time, link it to older knowledge, generate inferences and thus have its own representation of the world.

This knowledge is used to make decisions, or to interact with other agents or an external user.

AUTONOMOUS TRAIN

SpirOps is participating in the design of the operational center of the MARS LGV project (Autonomous Mobile Units for Security Reconnaissance), in consortium with SNCF Réseau, Forsee Power, Compagnie des Signaux, IRT Railenium, and Socofer France. The project is supported by BPI France as part of France 2030.

Learn more (article in french): "MARS LGV : une avancée technologique majeure pour la sécurité des Lignes à Grande Vitesse"

From 2018 to 2023, the team also carried out work on decision-making during obstacle detection as part of the Autonomous Train Passenger Service project, in consortium with SNCF Réseau, Railenium, Alstom, Bosch and Thales.

TOOLS

Since 2004, SpirOps has been developing its own data- and energy-efficient tools (C++, symbolic AI: explainable decisions requiring no training data, algorithmic optimization for real-time...).

SpirOps has also patented an artificial intelligence technology enabling complex behaviors to emerge from a multitude of simple rules.