Aurora's Road to Autonomy: Chris Urmson on Building the Future of Self-Driving

Explore the two-decade quest for autonomous driving with Aurora CEO Chris Urmson, a pioneer who co-founded Google's self-driving project (Waymo) and now leads the charge in driverless trucking. This article delves into the technological leaps, strategic decisions, and relentless perseverance required to bring self-driving vehicles from ambitious research to commercial reality. Discover insights on sensor technology, the business of autonomy, and Urmson's vision for a future where autonomous trucks transform logistics and eventually reshape personal transportation.

Key Takeaways

• Two Decades of Drive: Chris Urmson, a pioneer in autonomous vehicles, has spent 20 years pushing the boundaries, from DARPA challenges to co-founding Google's self-driving project (Waymo) and now leading Aurora.
• Trucking as the Frontier: Aurora focuses on autonomous semi-trucks, addressing the significant driver shortage in the logistics industry and improving efficiency and safety.
• Sensor Diversity is Key: Urmson emphasizes the critical need for multiple sensors (laser, radar, camera) for robust and safe autonomous systems, contrasting with single-sensor approaches.
• Beyond Driver Assistance: There's a fundamental distinction between driver assistance systems, which require human attention, and truly autonomous systems, which fully replace the driver.
• The "Chasm" of Autonomy: Bridging the gap from driver assistance to full autonomy requires a different technological approach, prioritizing both high recall and high precision to avoid false positives and missed events.
• Google's Incubation Advantage: The Google "black project" allowed the early self-driving car team to incubate and develop the technology with long-term vision, something challenging for a startup to achieve independently.
• Patience and Persistence: Building a self-driving company demands significant time and investment, as demonstrated by Waymo's 16-year journey to scale and Aurora's eight-year path to market.

Timeline Overview

• 00:00 Trailer: A sneak peek into the episode, highlighting Chris Urmson's vision for a self-sustaining freight business and the challenges and opportunities in the trucking industry.
• 00:43 Introduction: Host Jubin from Kleiner Perkins introduces Chris Urmson, co-founder and CEO of Aurora, discussing his background and the company's focus on autonomous semi-trucks.
• 01:59 FSD: are we there?: Chris Urmson discusses the current state of Full Self-Driving (FSD) technology, comparing Waymo's established presence to Aurora's upcoming autonomous trucking routes.
• 14:31 The competition, a million dollar check from LA to LV: Urmson recounts his experience with the DARPA Grand Challenge, where teams competed to drive autonomous vehicles across the desert for a million-dollar prize.
• 22:50 Dream like an amateur, execute like a pro: This segment delves into the origins of Google's self-driving car project, a "black project" designed to incubate the technology away from typical corporate bureaucracy.
• 32:30 Operate with integrity: Chris Urmson discusses the importance of integrity in business, reflecting on the challenging departure of a co-founder and the internal strife it caused within the Google self-driving project.
• 42:49 The future is here, unevenly distributed: Urmson talks about the rapid normalization of autonomous vehicles in areas where they operate and society's often underestimated adaptability to new technologies.
• 49:36 Underestimated decisions, minimizing regrets: Urmson reflects on significant decisions, including his departure from Google and the acquisition of Blackmore, a critical technology for Aurora's long-range sensing needs.
• 1:03:55 Retaining value: Chris Urmson discusses Aurora's strategy for maintaining its value in the public markets after its SPAC merger, navigating market volatility, and the "missionaries vs. mercenaries" dynamic among employees.
• 1:16:45 Integrating self-driving: Urmson details Aurora's approach to integrating its self-driving technology into new semi-trucks directly from the assembly line, focusing on partnerships with OEMs rather than retrofitting existing fleets.
• 1:28:20 Lifer: Chris Urmson expresses his long-term commitment to leading Aurora, driven by the company's unique combination of people, technology, partnerships, and business plan.
• 1:29:25 Who Aurora is hiring: Chris Urmson mentions that Aurora is continuously looking for talented engineers who are passionate about product delivery.
• 1:29:53 What “grit” means to Chris: Chris Urmson defines grit as "wrestling with it," "making it through it," and "hard work."
• 1:30:15 Outro: The closing remarks of the podcast episode.

The Decades-Long Pursuit of Autonomous Driving

Chris Urmson, a foundational figure in autonomous driving, has dedicated two decades to making self-driving vehicles a reality. His journey began at Carnegie Mellon, where he was the technical director for their DARPA Grand Challenge team, pushing robotic cars through the desert. This pioneering work led him to co-found Google's highly secretive self-driving car project, now known as Waymo. Today, as the CEO of Aurora, Urmson is spearheading the development of driverless trucking, believing the future of autonomy is "finally within reach."

Urmson points out that autonomy was always a matter of "when, not if." He highlights the immense societal benefits, such as saving lives, noting that "the US alone 42,000 people die on roads every year." This stark figure contrasts sharply with air traffic fatalities, underscoring the urgent need for safer transportation. Beyond safety, Urmson emphasizes the economic advantages, particularly for the trucking industry. The sector faces a significant challenge with a "90% turnover year" in drivers, making it "incredibly challenging" to build and sustain a business. Autonomous trucks can address this by providing a scalable, safe driver solution, operating assets "twice as hard" and dramatically reducing operational costs and improving fuel economy.

The "Chasm" Between Driver Assistance and True Autonomy

A critical distinction Urmson draws is between advanced driver assistance systems (ADAS) and fully autonomous driving. While systems like adaptive cruise control provide assistance, they still assume an attentive human driver. This allows for a "lack of recall" in the system, meaning it's okay if it occasionally misses an event, as the human can intervene. However, this approach prioritizes avoiding "false positives"—unnecessary braking or actions that frustrate drivers. "If you're driving your nice BMW or Mercedes with adaptive cruise control down the freeway at 60 mph or 65 and it hammers the brakes on in the middle of nowhere you're like 'That's not cool.'"

For a truly self-driving vehicle, where the human driver is no longer expected to pay attention, this "freedom" to have a lack of recall disappears. The system must achieve both "high recall and high precision," meaning it must see and react to all necessary events without errors. Urmson argues that relying solely on cameras, as some companies do, makes it incredibly difficult to "cross the chasm" to full autonomy safely and robustly.

The Imperative of Diverse Sensor Technology

Aurora's strategy, shared by Waymo, relies on a combination of different sensors: "laser, radar and camera." Each sensor type has its own strengths and weaknesses. By using multiple sensors, the system gains a comprehensive and redundant understanding of the environment, mitigating the risks of any single sensor being "fooled." This multi-sensor approach is particularly crucial for Aurora, which operates large trucks at highway speeds.

Urmson reveals that Aurora has developed a "special kind of laser rangefinder that allows us to see further than conventional LAR can," a critical innovation for enabling safe operation of 18-wheelers at 70 mph. This advanced sensor suite, combined with "high performance computing" and "AI and machine learning," allows their vehicles to "understand the world around it and drive safely through it."

The Birth of a Secret Project at Google

Urmson's journey to Google in 2009 was not a typical recruitment. Larry Page and Sergey Brin, Google's co-founders, had a vision for a "secret project" to build a self-driving car. This initiative, later named Project Chauffeur and eventually Waymo, was incubated as a "black project" to protect it from the "bureaucracy and stuff that happens for good reason" in a large company. This allowed the small, dedicated team to focus intensely on fundamental research and development without immediate pressure for commercialization.

Urmson recalls the intense work environment: "I was getting up and head to the office at 8 or 9... work till 6:00 go home have dinner put the kids to bed go back work... 8 9 till 1 in the morning." This dedication highlights the significant effort required to advance such complex technology. The long-term vision and resources provided by Google were crucial. Urmson believes that without Google's "willingness to invest and kind of explore what seemed like a crazy idea at the time," the autonomous vehicle industry would not be where it is today. He notes that Waymo's journey from 2009 to 2025—16 years to achieve scale—demonstrates the extended timeline required for such a transformative technology.

Dream Like an Amateur, Execute Like a Pro

Chris Urmson reflects on the early days of Google's self-driving project, admitting that "ignorance is bliss" was key. He suggests that if he and his team had truly understood "how hard this was going to be," they might have pursued other ventures. This aligns with the adage: "dream like an amateur and execute like a professional." It's a common thread for first-time founders, who often "do not appreciate what they're getting into." The complexity of real-world "edge cases" for autonomous driving, combined with the nascent state of applied AI at the time, made the task seem deceptively simpler than it proved to be.

Operate with Integrity: Navigating People Problems

While the technological hurdles of autonomous driving were immense, Urmson states that "it's always been people problems" that were the hardest to navigate. He explains that technology, given enough effort and smart people, will eventually be "figured out." However, interpersonal dynamics are "delicate." He speaks about prioritizing the success and well-being of his team, wanting their investment of time and effort to be "worth it for them."

A significant challenge for Urmson was the "An Lewendowski Google stuff" that generated considerable press. Anthony Levandowski, a co-founder of the self-driving project, was later convicted of stealing technology and taking it to Uber. Urmson recounts being "deposed" in the legal proceedings and the "stressful and difficult" environment this created within the team. He expresses regret for not being "effective enough at communicating" the internal strife to leadership, believing it "would have caused less pain for everyone involved." The incident served to solidify Aurora's core value: "operate with integrity."

Urmson also recounts a moment in 2011 when Levandowski attempted to convince core engineers, including Urmson himself, to leave Google and start an independent self-driving company. "I was like 'No I don't think that's right... I owe it to them to keep building this.'" This pivotal moment represented a real risk of the project being "torn apart," as these were "very important engineers" and the "core heart of the project." Despite the immense pressure and the "frustration," "fear," and "disappointment" he felt, Urmson and other key engineers chose to stay, a decision he credits with Waymo's eventual existence. He views Waymo's continued success, even after his departure, as a testament to the strong foundation and culture he helped establish.

The Future Is Here, Unevenly Distributed

Urmson notes the rapid normalization of autonomous vehicles in areas where they operate, like San Francisco. He describes how quickly people adapt to seeing driverless Waymos, transitioning from initial skepticism or awe to simply accepting them as "just part of the city." He attributes this to human adaptability and the familiarity of being driven by someone else. "My theory is that we are way more adaptable as humans than we think." This human trait, to "normalize incredibly quickly" to new situations, explains why technologies like the iPhone or services like Uber and Airbnb become commonplace so rapidly, despite initial doubts or even apprehension.

Underestimated Decisions, Minimizing Regrets

Urmson’s departure from Google in 2016 was a challenging decision, driven by disagreements over strategy and "people things." He believes in a philosophy of either helping fix a problem, getting in line, or getting "out of the way." Feeling misaligned with the company's trajectory, he chose the latter. This marked a significant personal shift, as he had previously envisioned himself as a "lifer" at Google.

After leaving, he spent several months reflecting on his next move, ultimately deciding to start Aurora. His motivation was to "minimize regret," not wanting to look back and wish he "could have been a contender" in the self-driving revolution. He recognized his "unique set of experiences and skills" in the autonomous vehicle space and felt a sense of "purpose" to continue contributing.

A crucial strategic decision for Aurora was the 2018 acquisition of Blackmore, a company specializing in frequency-modulated continuous wave (FMCW) LiDAR. Initially, Urmson and his team didn't believe traditional LiDAR could see far enough for safe high-speed trucking operations. Blackmore's FMCW technology offered a "10 to 20-fold amplification of the signal," allowing for much greater range with the same energy output. This acquisition proved to be "one of the best acquisitions you could imagine," providing a fundamental technological component that is "on every one of our trucks today."

Urmson chose to start Aurora as an independent company rather than joining an existing OEM or tech giant because he believes a "well-run company executes on its mission." He observed that other companies, like Apple, Ford, Volkswagen, and General Motors, eventually scaled back or shut down their autonomous vehicle efforts because their core missions were not solely focused on self-driving technology. By building Aurora as a dedicated entity, he aimed to "control our destiny" and ensure that the immense investment required in autonomous driving would remain aligned with its core mission.

Retaining Value

Aurora's journey to the public markets through a SPAC merger in 2021 was timed as the market for such ventures was booming. However, Urmson acknowledges the "rough for a while" period that followed, with the stock price plummeting significantly. He highlights the emotional toll of such a decline, particularly for investors whose "retirement" funds were affected, and for employees who saw the value of their equity diminish before their lockup periods ended. "You've just taken a chunk of some people's retirement and dropped it by 85% right that sucks."

Despite the challenging market conditions and being "thrown in with the rest of the stuff that's kind of crap," Urmson emphasizes the importance of integrity and communication with investors. He consistently reiterated that while the market was volatile, the underlying business and its long-term vision remained strong. "If you believe in what we're doing here if we keep our heads down and execute it'll come back." This honest approach fostered trust and helped Aurora retain value when many other SPACs did not.

A silver lining of this period was the "great filter" it created for employees, distinguishing "missionaries" (those committed to the vision and mission) from "mercenaries" (those primarily motivated by financial gain). While difficult to witness colleagues leave, Urmson noted that the remaining team members were "on the bus for the right reason," leading to a stronger, more focused culture. This cultural shift, where "the questions... are no longer about like hey stocks down what are you going to do about it and it's more hey how are you going to make the company succeed," was a "secret blessing."

Integrating Self-Driving

Aurora currently operates about 30 autonomous semi-trucks, with 10-15 running daily, accumulating over 2.5 million miles for customers like FedEx and Schneider. While they currently have an operator on board, Aurora aims to remove them as early as April. These operations are primarily concentrated on the "southern freight belt," starting with the Dallas-Houston corridor and expanding to Fort Worth-El Paso and Phoenix.

Urmson explains Aurora's strategy for integrating its self-driving technology. Rather than retrofitting existing trucks, they are partnering with OEMs like Volvo and PACCAR (maker of Peterbilt) to have the "Aurora driver" installed directly on new trucks coming off the assembly line. This approach aligns with the trucking industry's practice of frequently turning over their fleets due to high mileage, maintenance costs, and driver preferences for new vehicles. Trucking companies will purchase the truck from the OEM and have a separate subscription with Aurora for the autonomous driving service, paying on a "mile on a mile basis." This model is designed to "slot in" seamlessly with existing business operations and address the significant driver shortage.

Regarding the broader consumer market, Urmson anticipates that ride-hailing services will be the primary entry point for fully autonomous vehicles. He believes that as the cost of autonomous services decreases, they will eventually become more economically viable than personal car ownership, especially in dense urban areas, mirroring the shift seen in cities like New York. While initially replacing secondary cars in families, he envisions a future where urban dwellers might opt out of car ownership entirely, relying on ubiquitous autonomous ride-hailing services.

Lifer

Despite the challenges inherent in pioneering such a transformative technology, Chris Urmson expresses a strong and ongoing commitment to leading Aurora. He states, "forever is a long time but for the foreseeable future yeah," indicating his dedication to the company's long-term success. Urmson feels a unique connection to Aurora, believing he "wouldn't want to be the CEO of any other company," particularly not another AV company. He attributes this sentiment to Aurora's "combination of the people we have, the technology we have, the partnerships, the business plan that we have," which he finds "incredibly exciting."

Who Aurora is Hiring

Chris Urmson confirms that Aurora is "always hiring." He specifies that they are continuously looking for "engineers who are awesome" and possess a "passion for serving the product" and "delivering what we need together." While not highlighting any particular specific roles, the company seeks individuals who are committed to the practical application and successful deployment of their technology.

What “grit” means to Chris

When asked about the meaning of "grit," Chris Urmson offers a two-part response. His initial, slightly humorous thought is "dirt on the road," a nod to the practical, real-world challenges of autonomous driving. More profoundly, he defines grit as "wrestling with it," "making it through it," and simply "hard work." This definition encapsulates the persistent effort and resilience required to overcome the immense obstacles in the long and arduous journey of building self-driving technology.

Practical Implications of Autonomous Driving:

For Logistics & Trucking:

• Solves Driver Shortage: Autonomous trucks address the critical lack of human drivers.
• Boosts Efficiency: Trucks operate longer, increasing asset utilization and reducing costs (driver wages, fuel).
• Enhances Safety: AI-driven, multi-sensor systems significantly reduce accidents caused by human error.
• New Business Models: OEMs integrate self-driving tech into new trucks, offered as a subscription service.

For Consumers & Personal Transport:

• Shift to Ride-Hailing: Autonomous vehicles will first scale via services like Waymo, reducing the need for personal car ownership in cities.
• Urban Redesign: Less car ownership could free up city space for other uses.
• Driving as a Choice: For future generations, a driver's license may become optional.
• Increased Accessibility: Greater mobility for non-drivers (elderly, disabled).

Broader Impact:

• Economic Shift: Reshaping trillion-dollar industries, displacing some jobs while creating new tech and maintenance roles.
• Investment & Partnerships: Continued opportunities in AV tech and collaborations between AV companies and OEMs.
• Rapid Adaptation: Society will likely adjust to autonomous vehicles faster than anticipated.

Conclusion

Chris Urmson's two-decade quest for autonomous driving underscores the immense technological and societal challenges, yet also the profound benefits, of a driverless future. The journey from DARPA challenges to pioneering autonomous trucking with Aurora highlights the crucial need for diverse sensor technologies and patient, long-term investment.