What's Inside the Trunk of a Waymo Robotaxi?
Waymo has, for the first time, publicly shared details about the powerful computer sitting in the trunks of its self-driving taxis. The details, published on August 20, 2026, include chip architecture, processor specs, internal component details, and the names of the hardware suppliers Waymo relies on. The blog post was written by Satish Jeyachandran, VP of Engineering, and Daniel Rosenband, Compute Lead. (The Verge)
Waymo calls this onboard computer the "brain" of the Waymo Driver. It uses server-grade processors, the kind you would normally find in a data center, to process information from dozens of sensors mounted on the vehicle. Those sensors include cameras, lidar (a technology that uses laser pulses to measure distances and build a 3D picture of the surroundings), and radar. Waymo combines machine learning with other processors to balance the heavy work of understanding the road with the routine work of keeping the vehicle running. (Waymo, The Verge)
The company says its computer design follows three rules it will not compromise on: it must be responsive, ruggedized, and redundant. Responsive means it reacts fast enough to drive safely. Ruggedized means it can handle vibration, heat, and the daily wear of driving. Redundant means there are backups, so a single part failing does not cause a dangerous situation. The system must also run nearly silent and take up minimal trunk space so passengers do not notice it. Waymo designs its compute system in-house but works with outside suppliers for parts it cannot make itself. (The Verge)
The amount of computing power has grown a lot. Waymo says its onboard compute has increased 20 times over the past eight years, as the software that detects objects and plans the car's next move has become more advanced. (The Verge)
The disclosure comes as Waymo is operating at a large scale. The company runs roughly 4,000 vehicles across more than 10 cities, doing about 500,000 paid trips per week. The fleet drives more than 4 million miles every week. Its latest safety report covers more than 220 million fully autonomous miles through the end of March 2026, across five areas including Atlanta for the first time. (The Verge, Waymo)
The safety data stands out. Compared to human drivers in the same areas over the same period, the Waymo Driver was involved in 94% fewer crashes causing serious or fatal injuries, 82% fewer crashes where an airbag deployed, and 82% fewer crashes involving any reported injury. For people outside vehicles, the numbers are also strong: 93% fewer injury-causing crashes involving pedestrians, 84% fewer involving cyclists, and 84% fewer involving motorcyclists. In Atlanta alone, across more than 5.4 million autonomous miles, Waymo recorded 94% fewer airbag-deployment crashes and 86% fewer injury-involving crashes than the human benchmark, both statistically significant. (Waymo)
In everyday terms, Waymo estimates its performance equals one fewer serious-injury-or-worse crash every eight days, roughly six fewer airbag-deployment crashes per week, and about thirteen fewer injury crashes of any kind per week. Over its full operating history, the company estimates 47 fewer serious-or-fatal-injury crashes, 305 fewer airbag-deployment crashes, and 707 fewer injury crashes of any kind than if humans had driven those same miles in those same locations. (Waymo)
Carol Flannagan, a Research Professor at the University of Michigan Transportation Research Institute (UMTRI), noted that Waymo now drives enough miles to make fair comparisons to human drivers on crash rates and praised the consistency of results across locations. The full data and methodology are published at waymo.com/safety/impact. (Waymo)
The computer disclosure also makes the contrast with Tesla's approach clearer. Tesla CEO Elon Musk has called lidar a "crutch" and "a fool's errand," arguing that combining inputs from cameras, radar, and lidar creates "sensor contention," where the sensors disagree and the system has to sort out conflicting information. Waymo's approach is the opposite: it builds its system around combining all three sensor types, which requires a lot of computing power to do in real time. (The Verge)
The cost of all this is still high. Waymo's current robotaxi vehicles cost over $120,000 each, a figure Reuters reported could eventually fall to $85,000. The company took many years to build a fleet of 1,500 robotaxis and has also been conducting tests in Japan. (Reuters, Reuters)
The broader context here is that the decision to publish chip architecture, processor specs, and a supplier list is a real shift for a company that has historically kept its hardware details secret. For competitors and suppliers alike, the blog post gives the first concrete look at what is inside the trunk of a Waymo vehicle and the supply chain behind it. Whether that openness reflects confidence in a lasting hardware lead, a response to regulatory or partner pressure, or an effort to make the technology feel more normal to the public, the post does not say.
What it does make clear is that Waymo's computing challenge is enormous, clearly defined, and being handled at a scale few competitors can match. A 20-fold increase in computing power over eight years, combined with a sensor setup that processes dozens of inputs in real time across a 4,000-vehicle fleet driving 4 million miles weekly, puts Waymo in a category that is hard to copy without similar investment and discipline. The safety data, now covering 220 million miles across five areas with statistically significant results, adds a level of real-world evidence that the self-driving field has lacked for most of its history.
The computer design Waymo has revealed, the safety record it has published, and the scale at which it now operates are not separate stories. They are the same story: a system where the computer, the sensors, the vehicle, and the real-world deployment are all designed to work together. The trunk-mounted "brain" is what ties it all together, and for the first time, we can see what it is made of.
