Tesla vs Waymo: Miles Per Incident

The Numbers at a Glance

On raw miles per incident, Waymo leads by roughly 10x. But this snapshot comparison misses crucial context about trajectory, history, and what these numbers actually represent.

Context: Very Different Histories

Waymo traces its origins to the Google Self-Driving Car Project, which began in 2009. That is over 16 years of continuous development, testing, and real-world operation. Waymo launched its first commercial ride-hailing service in Phoenix in December 2018, giving it over 7 years of commercial operations.

Tesla launched its unsupervised robotaxi service in Austin in June 2025, less than a year ago. While Tesla has been developing its Full Self-Driving (FSD) software since 2020 and accumulating supervised driving data from millions of consumer vehicles, the transition to fully autonomous, unsupervised commercial service is a fundamentally different challenge.

Comparing Tesla's current MPI to Waymo's is like comparing a startup's first-year revenue to an established company's annual figures. The absolute number matters, but the growth rate matters more for predicting the future.

Tesla's Unprecedented Improvement Rate

What sets Tesla apart is the speed of improvement. Our tracker shows that Tesla's miles per incident has been doubling approximately every 79 days. This is an exponential improvement trajectory that, if sustained, has dramatic implications:

• At 79-day doubling: Tesla could match Waymo's current MPI within 3-4 more doublings (roughly 6-9 months)
• Each doubling represents a 50% reduction in incident rate
• The improvement is driven by over-the-air software updates deployed across the entire fleet simultaneously

No other autonomous vehicle company has publicly demonstrated this rate of safety improvement. Waymo's safety metrics have been consistently high but are not known to be improving at this exponential rate, likely because Waymo is already operating at a very high baseline.

There is no guarantee that Tesla can sustain this trajectory. Exponential improvement may slow as the "easy" problems are solved and remaining edge cases become harder. But even a slowdown from 79-day doubling to 150-day doubling would still represent rapid progress.

Different Reporting, Different Baselines

Direct MPI comparisons between Tesla and Waymo are complicated by differences in how safety data is reported and measured:

Data Sources

• Tesla's MPI on this tracker is calculated from NHTSA SGO 2021-01 incident reports and independent fleet tracking. The data is public and verifiable.
• Waymo's MPI comes primarily from the Swiss Re/Waymo study, which used insurance claims data. Waymo also files SGO reports, but their published safety figures rely on their own methodology.

Reporting Thresholds

SGO reports capture incidents at a lower severity threshold than insurance claims. This means Tesla's NHTSA-reported MPI may capture minor events that would not appear in Waymo's Swiss Re-based figures. If both companies were measured using identical criteria, the gap might be narrower (or wider) than the raw numbers suggest.

Geographic Differences

Tesla currently operates unsupervised only in Austin, Texas. Waymo operates across Phoenix, San Francisco, Los Angeles, Austin, and Miami. Different cities present different driving challenges. Austin's road environment may be easier or harder than Waymo's multi-city average, making geographic comparisons imprecise.

Different Fleet Sizes and Approaches

The two companies take fundamentally different technical approaches:

• Waymo uses a multi-sensor suite including LiDAR, cameras, and radar on purpose-built Jaguar I-PACE vehicles. Each vehicle costs significantly more to produce and maintain. Waymo's fleet is estimated at roughly 700+ vehicles across all cities.
• Tesla uses a vision-only approach (cameras only, no LiDAR) on modified Model 3 and Model Y vehicles. This approach is cheaper per vehicle and is designed to scale to Tesla's existing production capacity. Tesla's Austin fleet is currently around 80 unsupervised vehicles.

Tesla's vision-only bet is more ambitious and more cost-efficient if it works at scale. Waymo's sensor-rich approach provides redundancy and has proven effective at achieving very high safety levels. Both approaches have trade-offs, and the market may ultimately support both.

Both Are Making Roads Safer

It is easy to frame this as a competition, but the bigger picture is that both Tesla and Waymo are demonstrating that autonomous vehicles can operate safely on public roads. Neither company has had a fatality in unsupervised autonomous mode. Both have incident rates that are improving over time.

Human drivers in the United States cause approximately 40,000 deaths per year and over 5 million reported crashes. Any autonomous vehicle system that achieves better safety than the human average, even in limited conditions, represents progress toward reducing this toll.

The competition between Tesla and Waymo is likely to accelerate safety improvements for both. Tesla's rapid improvement may push Waymo to continue innovating. Waymo's established safety record provides a benchmark for Tesla to target.

For riders and the public, the relevant question is not "which company is winning?" but "are both getting safer over time?" The data suggests yes. Track Tesla's progress on our main safety tracker and explore how the fleet is expanding across multiple cities.

What to Watch Going Forward

Key indicators that will shape the Tesla vs Waymo comparison in 2026 and beyond:

• Tesla's doubling time: If the ~79-day doubling rate holds, Tesla could approach Waymo-level safety by late 2026. If it slows significantly, the gap may persist.
• Tesla's expansion: As Tesla launches in new cities (Dallas, Houston, Phoenix planned for H1 2026), maintaining safety performance in new environments will be a crucial test.
• Waymo's scaling: Waymo is also expanding aggressively. Whether they can maintain their safety record at larger scale is an open question.
• Monitor removal: Tesla's ongoing removal of safety monitors in Austin will reveal whether the safety metrics hold without human backup.
• Standardized reporting: If NHTSA or the industry develops more standardized safety metrics, direct comparisons will become more meaningful.