LeddarTech and TTTech Auto have announced a significant milestone in autonomous driving safety with the completion of a joint demonstrator for fail-operational architecture. LeddarTech, an automotive software company specializing in AI-based low-level sensor fusion and perception software technology, collaborated with TTTech Auto, a leader in automotive safety software, to develop a fail-operational architecture based on LeddarTech’s LeddarVision software and TTTech Auto’s Safety Co-Pilot solution.
TTTech Auto’s Safety Co-Pilot solution is designed to address the challenges of a safety architecture suitable for autonomous driving beyond SAE Level 3. It consists of a doer channel responsible for the primary driving task and an independent checker channel. In the event of an abnormal situation identified by the checker, the doer channel is overridden, and a fallback channel safely brings the vehicle to a stop.
The collaboration between TTTech Auto and LeddarTech combines TTTech Auto’s expertise in safety mechanisms and automotive platforms with LeddarTech’s advanced fusion and perception capabilities. TTTech Auto contributes operational platforms and frameworks, facilitating the development of decomposed architectures for safety applications, while LeddarTech enhances the partnership with its high-performance fusion and perception stack. Both companies adopt an agile approach, employing risk assessment and safety-critical event feasibility assessments to demonstrate a potential solution for car OEMs constructing their SAE Level 2-3 safety solutions.
To showcase the viability of their approach and its potential for vehicle safety and cost-efficiency, TTTech Auto and LeddarTech have created a joint demonstrator focused on the checker channel. The key features include LeddarTech’s LeddarVision, an AI-based low-level fusion and perception software stack, and TTTech Auto’s Safety Co-Pilot, a versatile solution equipped with validation modules for critical safety requirements. The checks are performed based on a separate environment model, LeddarVision, ensuring independence from the doer channel.
The demonstrator was tested using data collected from highways in Germany and Israel. The results validate the efficiency and reliability of LeddarTech’s environmental model and TTTech Auto’s probabilistic checking in the checker channel. This advancement provides a robust safety argument for autonomous driving systems operating at SAE Level 3 and beyond.
Stefan Poledna, CTO at TTTech Auto, expressed their commitment to fail-operational architectures, saying, We are paving the way to autonomous driving safety with true fail-operational architectures. Frantz Saintellemy, President, COO, and incoming CEO of LeddarTech, emphasized the importance of partnerships and collaboration in accelerating technology deployment and adoption.
TTTech Auto provides solutions for future vehicle generations and the shift towards software-defined vehicles, while LeddarTech develops comprehensive AI-based sensor fusion and perception software solutions for ADAS and autonomous driving applications. Both companies are dedicated to improving safety and efficiency in global mobility.
This joint development between LeddarTech and TTTech Auto signifies a significant step forward in autonomous driving safety, bringing us closer to the realization of fail-operational architectures and safer autonomous vehicles on the road.
