On their way to building fully self-driving vehicles, self-driving automakers face a daunting task: training their AI systems to be able to reliably respond to any and all scenarios a car, truck, bus, or, hopefully, vehicle might encounter. Better than a human could do. Today, a startup with a platform to help address this challenge is announcing a major round of funding to advance these strategies.
Foretellex, which builds verification and validation solutions to test the full range of driver assistance and autonomous systems coming to market, closed its Series C at $85 million. The round includes financial investors alongside strategic backers from the automotive and chip industries, and is an indication of who is already engaging with Foretellix, as well as the startup’s longer business trajectory.
The round is led by Israeli VC firm 83North, with Singaporean Temasek and automaker Isuzu investing alongside Woven Capital (Toyota’s investment fund), Nvidia, Artofin, and previous backers MoreTech, Nationwide, Volvo Group VC, Jump Capital, Next Gear Ventures, and OrCrowd. . . Foretellix may ring a bell for readers: The first closing of this round was in May of this year ($43 million).
This brings the total amount raised by Foretellix to $135 million. He did not disclose the valuation but co-founder and CEO Ziv Binyamini confirmed to TechCrunch that it is higher than the previous figure, but “not crazy.”
The company’s client list includes some of the biggest names in the automotive industry. These companies include Volvo, Daimler, Isuzu, Toyota, “and other companies that have not yet gone public,” Benjamini said.
At previous companies including Intel and Cadence Design Systems, Benjamini and his co-founders Gil Amed and Yoav Hollander developed and refined much of the knowledge about testing and computation as a means of exploring performance and design across a number of variations, permutations, and uses. But specifically with regard to chip design. It was about “finding the problems before they committed to them [manufacturing] “Silicon,” he says.
In Benjamini’s view, scenarios in automotive and autonomous systems are another step along this path, albeit a more complex and dynamic one.
There is some artificial intelligence playing a role in how Foretellix does its work, and its goal is partly to improve AI systems, but in reality the core of their breakthrough was to come up with the algorithms, and computations, needed to create many scenarios (expressed in the form of data and images) to help train the systems. on what might happen, and make sure that systems know how to respond to it (and that they respond at all).
He describes the core of the solution stack as the ability to “automatically generate scenarios – an unlimited number of scenarios – based on high-level specifications” in the language that has now become the industry standard. (He refers to the standard development of ASAM OpenSCENARIO 2.0 (OSC2.0), which was developed in part by Foretellix itself to integrate across the highly fragmented and proprietary landscape of self-help and driver platforms and systems.) “To find edge cases in millions and millions of different situations,” he said. Besides, there is a large big data analysis component, which analyzes test results to give customers deeper insights.
Some of the funding will go towards further development of the platform – including bringing in more productive AI to operate the system as well as to interact with it – as well as to develop the business. Today the company employs about 180 employees, and the plan is to add more.
There’s a very clear market segment in consumer motors, but likewise the company is building solutions for a range of other vehicles, from trucks and buses to vehicles used in enclosed campus parking – for example, in industrial scenarios like mining – where full autonomy already exists. .
This has also included contributing to other companies’ simulation testing platforms. Nvidia, for example, has Integrated Foretellix tools In the Drive SIM driving simulation platform, used to test autonomous systems.
“Isuzu believes that safety verification and validation is essential for OEMs to responsibly develop safe autonomous vehicles,” Hiroshi Sato, vice president of engineering at Isuzu, said in a statement. “Foretellix’s advanced technology in creating automated scalable scenarios, and its leadership position in the OpenSCENARIO 2.0 standard, are a major asset for Isuzu. Collaborating with Foretellix’s senior management and engineers is a great opportunity and a great advantage for the development of Isuzu’s autonomous vehicles.”