California’s Division of Motor Vehicles has approved Nuro to test its third-generation R3 autonomous delivery vehicle in four Bay Area cities, marking a significant milestone for the struggling startup.
Nuro has been granted approval to test its autonomous delivery vehicle in four California cities: Mountain View, Palo Alto, Los Altos, and Menlo Park. Nuro’s autonomous vehicles, devoid of seats, windows, steering wheels, and pedals, are specifically engineered for transporting goods alone. Despite traversing public roads, these vehicles appear as oversized sidewalk service robots, equipped with climate-controlled storage compartments designed to transport food.
The revamped geographical area will host the third-largest autonomous vehicle deployment in the US, potentially rivalling Waymo’s scale, according to Cruise co-founder Dave Ferguson, who noted that the company may have required more time for its larger-scale rollout before temporarily grounding its fleet last year.
Nuro, meanwhile, has been operating under a 10-year agreement with Uber Eats to pilot its autonomous delivery service using third-party vehicles.
Nuro, which has been operational for a few years, made the decision last year to put on hold its planned manufacturing push with Chinese electric vehicle manufacturer BYD, a collaboration that could have resulted in the production of hundreds of vehicles. The startup, which had once been hailed as a darling of the AV industry after securing more than $2 billion in funding, was hemorrhaging cash at an alarming rate. Following successive rounds of layoffs over the past couple of years, the company is now focusing on giving its remaining workforce the autonomy they need to thrive. The company was forced to put its automotive manufacturing and business operations on hold temporarily.
Ferguson tells TechCrunch that Nuro is not planning a swift resumption of large-scale manufacturing and commercial operations. The corporation remains intensely focused on testing and validating its novel artificial intelligence architecture, with Ferguson noting that this approach is already yielding tangible benefits.
According to Ferguson, the company has significantly hastened its advancements in autonomy, extending this momentum to various facets of the project as well. “In this Level 4 autonomous system, the software program is deeply intertwined with hardware, sensing technologies, and computing capabilities.”
The Society of Automotive Engineers (SAE) defines Degree 4 autonomy as the capability to drive itself without human intervention in specific situations, thereby enabling vehicles to operate independently under certain conditions.
Ferguson noted that Nuro has been rigorously testing and validating the R3’s innovative hardware and software ecosystem on a fleet of retrofitted Toyota Priuses – approximately 100 vehicles, according to an individual familiar with the matter – and has also continued some deliveries using these test cars for Uber Eats. In 2022, we embarked on a decade-long strategic alliance, marking the beginning of a fruitful 10-year business partnership.
Despite BYD’s decision to halt production, Nuro successfully secured around two dozen R3s from the electric vehicle manufacturer. In the coming months, Nuro plans to deploy its autonomous vehicle fleet within the San Francisco Bay Area and expand into its existing market in Houston.
Uber’s representative informed TechCrunch that they anticipate commencing use of the R3 for delivery purposes this autumn.
“One key advantage of the R3 over the R2 is its potential to operate in a significantly broader operational design domain,” said Ferguson. The R2’s top speed is a modest 25 miles per hour. The R3 will be able to drive up steep inclines at a pace of approximately 45 miles per hour. “We won’t necessarily roll out the technology at that pace initially, but it enables us to conduct comprehensive L4 autonomous testing, deployment and commercialization across a much broader area – essentially everything except highways.”
AI advancements at both the firm and industry levels have enabled Nuro to make significant progress. Ferguson noted that Nuro’s approach has evolved to leverage one or two large-scale foundational AI models that handle numerous tasks, including mapping, localization, comprehension, prediction, and planning, thereby yielding enhanced efficiency and effectiveness. Nuro integrates this functionality with a conventional system, where all tasks are executed autonomously by AI models, thereby validating its AI in real-time.
Nuro’s R3 deployment in the Bay Area and Houston signifies its ability to operate at a faster pace and cover larger territories, thereby setting the stage for future scaling.
Given the circumstances won’t recur this year, Nuro is unlikely to produce its self-driving vehicles using BYD’s manufacturing capabilities, as any products resulting from their partnership will likely be subject to stringent regulations. While Ferguson acknowledged tariffs as a potential issue, he expressed confidence in BYD as a reliable production partner.
As it continues operations, Nuro will remain focused on honing its skills and maximizing the value extracted from each Uber Eats delivery. Ferguson noted that Nuro is investigating alternative pathways to market beyond autonomous logistics, but refrained from providing further details.
