VALEO GROUP
13 October 2026
5 min
Valeo Group
Valeo and the open science lab KE:SAI will jointly develop and openly release models, code and weights for autonomous driving research.
October 13, 2026 — Paris, France — Valeo, a global leader in automotive technology, and KE:SAI (Kyutai ELLIS Scalable Autonomous Intelligence), the non-profit open science research lab co-founded by kyutai and the ELLIS Institute Tübingen, announce an open-science research collaboration to develop a fully reproducible, generalist AI driving system for autonomous vehicles.
The program aims to first match, and then push beyond, state-of-the-art performance across all major public driving benchmarks, and to release every component (code, trained models and datasets) as open source to build a strong foundation for democratizing self-driving technology and physical AI. Findings will be shared openly through publications and suppliers and researchers will be able to reproduce the results, build on the model and benchmark their own systems against a common reference.
Marc Vrecko, CEO of Valeo Brain, commented: “This open-science collaboration with KE:SAI marks a strategic milestone for Valeo and the broader European automotive ecosystem. By combining our deep industrial and technological expertise in autonomous driving with world-class open-source research, we are establishing a shared foundation to accelerate the development of physical AI and democratize next-generation self-driving technologies.”
Andreas Geiger, CEO of KE:SAI, commented: “For language models, open source compressed years of progress into months, yet frontier physical AI is still largely closed. Valeo shares our belief that open foundations create more innovation than any closed system ever could. Everything we develop will be open. Every step forward is a step forward for everyone.”
A Franco-German Initiative
The partnership brings together two of Europe’s leading teams in autonomous driving research, in Paris and Tübingen: valeo.ai, Valeo’s AI research lab, whose open-source models and publications, such as DrivoR and VaVAM, are widely used in the community, and the KE:SAI founding team, known for their open benchmarks and driving stacks such as KITTI, NAVSIM and TransFuser. The partnership’s ambition is to strengthen the European and worldwide automotive ecosystems by giving automakers, suppliers and research labs a built in Europe, open and shared foundation for autonomous driving systems.
An Open Ecosystem for Self-Driving
Physical AI is widely considered the next frontier of artificial intelligence, but leading efforts, including those in self-driving, remain largely closed with nearly every driving model deployed in the automotive industry proprietary. Automakers and suppliers have no shared, open foundation on which to build, compare and improve autonomous driving systems, leaving them dependent on technology they cannot inspect, adapt or reproduce. At the same time, important research questions regarding efficiency and generalization remain open. Finding out which approaches will make physical AI work reliably at scale requires exploring a wide range of ideas which can’t be pursued by a single lab alone. Developing in the open lets researchers across institutions test, challenge and build on the work. Importantly, building the foundations together, this effort brings academic and industrial research closer together, enabling rapid progress everybody can participate in. This partnership aims to push open source autonomous driving research towards the frontier as we have seen players achieve in the LLM space.
The Technology
The partnership aims to build one compact AI model that can drive different vehicles, with different sensors, in different environments. The teams will begin by training on large-scale real-world driving recordings, with a model that is designed to transfer across different sensor and camera configurations. From there, the model will go beyond what recorded driving alone can teach. Through reinforcement learning in simulation and learned world models, it will be exposed to rare and safety-critical situations that are scarce or absent in real-world data.
About Valeo
Valeo is a leading global technology company creating innovative solutions and systems for automotive and technology partners worldwide. Valeo is structured around its POWER, BRAIN and LIGHT Divisions, and Valeo Service, the augmented service partner, for the aftermarket and new forms of mobility.
Valeo is committed to making mobility safer, more sustainable and affordable to all. The Group is playing a vital role in shaping the Car of Tomorrow: which will be electrified, safer and software-defined. Valeo is leveraging its global industrial footprint and its technological leadership in electrification, advanced driver assistance systems, lighting and software to capture an increasing share of value per vehicle.
Valeo is listed on the Paris Stock Exchange.
Valeo in brief: €20.9 billion in sales in 2025 | 100,000 employees worldwide | 29 countries | 149 production plants | 59 R&D centers | 19 distribution platforms (as of February 26, 2026).
Learn more at www.valeo.com
Find all our press releases: https://www.valeo.com/en/press-releases/
Media Contacts
Dora Khosrof | +33 7 61 52 82 75
Florentina Deca | +33 6 32 13 69 15
Caroline De Gezelle | + 33 7 62 44 17 85
Clémence Le-Roux | + 33 6 12 18 26 00
press-contact.mailbox@valeo.com
Investor Relations
valeo.corporateaccess.mailbox@valeo.com
About KE:SAI
KE:SAI — Kyutai ELLIS Scalable Autonomous Intelligence — is a non-profit frontier AI research lab co-founded by kyutai and the ELLIS Institute Tübingen in May 2026 with a founding donation from kyutai. Co-located in Tübingen and Paris, this Franco-German collaboration merges kyutai’s open science ethos and foundation model expertise with ELLIS’s talent attraction and foundational research excellence in 3D computer vision, data-driven simulation, causality and physical AI to pioneer world-leading research in world models and autonomy. KE:SAI’s goal is to conduct fundamental research in physical AI and to train world models and driving policies that generalize effectively across a wide variety of embodiments and environments. KE:SAI has been co-founded by Andreas Geiger, Bernhard Schölkopf, Kashyap Chitta, Daniel Dauner and Bernhard Jaeger, a team of world-leading researchers in computer vision, robotics and machine learning.
Learn more at: https://kesai.eu/
