21-05-2025
Foretellix Unveils New AI Platform To Accelerate Autonomous Vehicles
Synthetic traffic scenarios can be created in realistic detail so that a comprehensive driving ... More environment is presented to an Autonomous Driving system.
Foretellix, a provider of data-automation for AI-powered autonomy, today announced a significant expansion of its Foretify platform, which the company says will enable Autonomous Vehicle (AV) developers 'to slash development time in half and save hundreds of millions of dollars.'
Foretellix views the emergence of the physical AI frontier presenting AV developers with opportunities for significant industry advances beyond the manual, rule-based software predominantly deployed for assisted driving capabilities. The company emphasizes that AV developers need a way to train the AI engines by maximizing the value of their existing drive data and enhancing it with hyper-realistic synthetic data. However, they also face serious challenges in preparing for the unpredictability of the technology. The need to validate that the AV stack is inherently safe in any possible scenario is vital for autonomy to be successfully launched on public roads.
Foretellix developed its Foretify data automation toolchain to train AI-powered AV stacks by automatically curating training data from the real-world driven miles and augmenting them with generated synthetic scenarios. The toolchain also provides diverse sensor simulation data enhanced with hyper-realistic variations generated by NVIDIA Omniverse and Cosmos. The AI-powered AV stack is validated by generating massive amounts of realistic synthetic data, increasing efficiency in reaching an optimal level of operational safety in a particular Operational Design Domain by tenfold, the company says. Furthermore, Foretify evaluates the real-world and simulation data coverage and provides evidence for the system safety case.
Backed by leading investors including Toyota's Woven Capital, Temasek, Volvo and NVIDIA, Foretellix was founded by pioneers of data-driven verification and validation tools for complex systems. The founding team's work across coverage-driven methodologies, verification, and validation at Verisity, Intel, and Cadence Design Systems is now widely used in the semiconductor industry and formed the basis of Foretify. Foretellix customers include Daimler Truck subsidiary Torc, Volvo, Mazda, Woven by Toyota, and Nuro.
'Physical AI creates a fast lane to autonomy that will ultimately improve our quality of life and help save millions of lives, but making it a reality requires an intelligent, data-driven approach to address AI's inherent limitations,' said Ziv Binyamini, CEO and Co-Founder of Foretellix. 'Training, validating and providing evidence of autonomous systems' safety are the key drivers for this industry to progress. Foretellix's solution helps foster innovation and reduce the barriers to entry for AV developers, vehicle manufacturers and entrepreneurs, democratizing the road to driverless autonomy.'
"Foretellix is a valued partner in our mission to commercialize autonomous trucks by 2027' said Torc CEO Peter Vaughan Schmidt. 'Their Data-Automation Toolchain enables the large-scale simulation and rigorous scenario testing essential to validating our physical AI level-4 autonomous system and ensuring that safety remains at the core of everything we do."
"By leveraging NVIDIA Omniverse and NVIDIA Cosmos, the Foretify toolchain can evaluate and generate high-fidelity testing scenarios and training data at scale," said Zvi Greenstein, General Manager, Autonomous Vehicles Infrastructure at NVIDIA. 'Foretellix's data-automation toolchain delivers the training, validation and safety evaluation data required to safely deploy AI-powered autonomy'
Foretify recently expanded its simulation capabilities through its integration with the NVIDIA Omniverse Blueprint for Autonomous Vehicle (AV) Simulation and the NVIDIA Cosmos Transfer World Foundation Model. The addition of hyper-realistic sensor simulation to Foretify's behavioral scenario simulation accelerates AI-powered AV training and validation for faster development and deployment.
The AI community continues to reach ever more deeply into creating a hyper-realistic synthetic world. This means safety teams at AV developers can throw a vast set of road driving scenarios at their self-driving system. This basic concept has been cooking for close to a decade. With each iteration, complexity and thoroughness seem to be growing by orders of magnitude.
