Waabi says its approach is different from other companies because of its focus on artificial intelligence. The Waabi Driver uses an AI-first autonomy stack, according to the company, allowing it to learn from an advanced driving simulator the company developed.
The Waabi Driver has two components: an advanced virtual driver as software, along with autonomous sensors and computer hardware. Together, they form a solution designed for factory-level OEM integration, large-scale commercialization, and safe deployment, according to the company.
“What makes Waabi Driver so unique is that it is purpose built for factory-level OEM integration as well as our groundbreaking technology to solve self-driving at scale,” the company said in a blog post making the announcement.
“The Waabi Driver is powered by a revolutionary AI-first approach that is end-to-end trainable, interpretable, has superior generalization capabilities, and can adapt to various hardware configurations, unlocking new autonomous lanes and truck platforms with unprecedented speed. Furthermore, the hardware is designed with production intent, it is plug-and-play for easy installation, maintenance and servicing.”
An 'AI-First' Approach to Autonomous Trucks
The Waabi Driver is an end-to-end trainable system that automatically learns from data, speeding up development and enabling it to learn the complex decision-making needed for operating on the road safely, according to the company. Waabi contends that this contrasts the traditional approach “that is brittle, overly complex, and requires painstaking manual code adjustments.”
This AI-first approach is empowered by Waabi World, a self-driving simulator of sorts, which exposes the Waabi Driver to the diversity of scenarios needed to hone its driving skills, including common situations and safety-critical edge cases. It would otherwise take thousands of self-driving vehicles driving millions of miles each to experience these situations on the road, according to Waabi.
“This approach drastically reduces the need to drive in the real world, resulting in a solution that is much more sustainable, and both smarter and safer before the wheels even start turning,” according to the announcement. “On-road driving is primarily reserved for the final step of development: validation and verification. This is a major paradigm shift.”
Waabi said its system “boasts superior generalization capabilities so it can safely apply learned skills to unseen scenarios and brand new geographies, without ever having driven there before,” which would allow it to unlock new autonomous-trucking routes faster.
The company said the Waabi Driver is designed to be adaptable to multiple redundant truck platforms and easily integrated on the truck assembly line. The hardware is plug-and-play, lightweight, simple to maintain, and aerodynamic to maximize fuel savings, according to the announcement.
It uses multiple sensors such as LiDARs, cameras, and radars. Using the sensor simulation capabilities within Waabi World, the company can streamline sensor selection, integrate the latest technology quickly, and deliver flexible solutions for OEM partners.
Who is Waabi?
Waabi is new Canadian-based autonomous vehicle startup. It was founded by Raquel Urtasun, an expert in machine learning and computer vision, making it one of the few women-led AV startups. In 2017, Uber tapped Urtasun to lead its new branch of its Advanced Technologies Group when she was an associate professor of computer sciences at the University of Toronto.
Waabi came out of stealth in June 2021 having raised $83.5 million, among the largest Series A rounds ever raised in Canada.
The company has brought on a team of executives and engineers with experience in autonomous vehicle development and in trucking:
- Vivian Sun joined Waabi last year as its first chief commercial officer. She spent more than five years previously at TuSimple, where she was VP of business development.
- A team of engineers was brought on board to integrate sensors, lidar, vision and computing systems into trucks, led by Eyal Cohen, who was previously with Uber ATG, Otto, and Apple’s vehicle program. It also includes Jorah Wyer, who worked for robot truck startup Ike, Uber ATG and Apple; and J.D. Wagner and Paul Spriesterbach, who both previously worked for autonomous tech developer Aurora.
- In September, Waabi announced it was hiring Dustin Koehl as Head of Transportation. He's a 16-year trucking industry veteran, most recently with U.S. Xpress. He spent 12 years with Total Transportation of MS, where he was named an HDT Emerging Leader in 2016.
Originally posted on Trucking Info