Waymo Robotaxi Expansion Adds 250 Square Miles in Five US Cities

A bigger map, not a new city

Waymo is spending this week making the cities it already serves bigger. On October 27 the company said it would add nearly 250 square miles of driverless coverage over the following week, spread across Los Angeles, Nashville, Orlando, San Diego and the San Francisco Bay Area.18 No new market is part of the rollout. Each addition widens a zone where cars already run with nobody behind the wheel.4

Here is what changes in each city, according to the reports:

• Los Angeles: service now follows the coast into the South Bay, through Manhattan Beach and Redondo Beach to Torrance.110
• Nashville: the zone moves farther south, far enough to reach the zoo.3
• Orlando: coverage on the east side nearly doubles and downtown service gets deeper.1
• San Diego: the area runs north to La Jolla and takes in University City and Clairemont.5
• Bay Area: coverage pushes further into the South Bay, adds a link to BART at Berryessa station in San Jose, and now stretches from San Francisco down the Peninsula to San Jose and Campbell.1

Orlando is the one market with a published figure. It gains 53 square miles, bringing it to 120 square miles in total.47 Waymo did not give a per-city breakdown elsewhere and did not say how many vehicles will serve the new territory.16 That second gap matters. A bigger zone without more cars could mean longer waits, so the expansion's real effect on riders is still unknown.

Why the fill-in strategy matters

The coverage agrees on the facts and mostly agrees on what they mean. Several outlets argue that widening existing zones does more for riders than opening a new city, because every added neighborhood removes another trip that would otherwise need a human-driven ride.47 That reading holds up. A robotaxi zone that stops short of where people actually go keeps the service a novelty. The Berryessa BART connection shows the goal clearly: the robotaxi becomes one leg of a commute rather than a destination in itself.47

The pace in the newest markets is also notable. Nashville opened to the public in April with about 60 square miles. San Diego opened on September 1 with roughly 50 square miles that barely reached north of I-8, so getting to La Jolla within two months is a fast jump.15 October was already busy for Waymo before this announcement. It opened public service in San Antonio, San Diego and Tampa on October 14 and began pickups at Las Vegas's Harry Reid International Airport the following week.27 It also said it is coming to Toronto, with first riders targeted for 2028, and that it is laying groundwork in the Madrid region of Spain.8

The money and fleet plans point the same way. Waymo recently closed a $5 billion term loan, its first debt financing, which it called a step toward becoming a scaling commercial business.8 Reports say it plans to grow its fleet roughly fivefold to 20,000 robotaxis by 2027.9 Read together, those moves look less like a technology demonstration and more like a company building out a utility.

One point does not line up across the coverage. Electrek says Waymo now operates in 15 US metros and delivers more than 500,000 paid rides a week.1 A separate piece citing the Associated Press puts it at more than 400,000 weekly trips across six metro areas.12 The smaller figure looks out of date, given how many cities Waymo has launched this year.

Tesla's competing maps

The Tesla comparison runs through almost all of the coverage. Waymo announced its expansion a day after Tesla significantly widened its Robotaxi zone in Houston.2 Electrek points out that the nearly 250 square miles Waymo is adding roughly equals Tesla's entire expanded Houston zone of about 254 square miles.1

The two fleets look very different in size. Robotaxi Tracker counted just 34 Tesla vehicles in Houston.1 Other reports describe Tesla's combined Cybercab and Model Y robotaxi count in Texas passing 1,000 vehicles.14 A state registry reportedly listed 739 Tesla autonomous vehicles around October 9, but that counts registrations, not cars actually driving without a person on board.17 Electrek also reports roughly 45 Cybercabs shipped to a new Houston hub.23 Our reading: Tesla registers and stages more vehicles than it runs without supervision, and it does not publish the unsupervised share.17

Mileage is the clearest comparison. Waymo averaged 3.9 million rider-only miles a week in the second quarter. Tesla's driverless fleet logged 558,000 miles in its latest reported week, mostly in Austin, about one-seventh of Waymo's figure.15 In the week ending October 18, Tesla reported 438,000 miles with no safety monitor on board, up 30% from the week before.4 Tesla's cumulative figures are rising quickly: nearly 2 million unsupervised miles by the end of September and 3.1 million by October 18.17 CleanTechnica cites a 3.3 million figure from earnings day.16 The growth is real, but it starts from a much smaller base. Electrek notes that all of this landed on the same day Elon Musk said the race against Waymo "was never a real contest."1

FSD's timeline slips again

Tesla's consumer software news makes the gap clearer. On its Q3 2026 earnings call, AI chief Ashok Elluswamy said the first customer-owned cars could run unsupervised FSD around Q2 2027.1118 The offer is narrower than earlier promises. Owners would add their cars to Tesla's managed Robotaxi fleet, inside Tesla's existing zones. They would not get a car that drives itself anywhere.1119 The software would arrive with FSD v15, which has not yet shipped to customers.1718 Reports differ on tone. CleanTechnica framed the quarter as FSD and Robotaxi "finally taking off," while also noting the unsupervised target has slipped many times.16 Electrek called it the latest date on a promise nearly a decade old.11 Both accounts are fair, and they describe different things: FSD is selling well, and it is still not autonomous.

The sales numbers are strong. Tesla reports 1.74 million active FSD subscriptions, and about 75% of new North American deliveries included FSD.1617 The latest supervised builds are also improving. FSD v14.3.13 claims 20% faster reaction time from a rewritten AI compiler and better handling of emergency vehicles and school buses. A "Lite" version brings what Tesla learned on newer HW4 cars to older HW3 hardware.1420 Even so, all of these builds are still "Supervised." In California, every Tesla robotaxi still carries a safety driver, because Tesla has not gone through the state's process for driverless operation.1117

Regulators turn to rewriting the rules

Federal regulators were busy in the same week. On October 23, NHTSA proposed optional cyan marker lamps that would light up only while an automated driving system is in control and no one on board is expected to watch the road.2429 Because the lamps are tied to that condition, a car running a supervised system such as Tesla's FSD could never turn them on.24 The same proposal removes turn-signal, hazard and beam-switch control requirements for vehicles built without manual controls.27 Comments are due November 25.2425

Purpose-built vehicles benefit most. Zoox needed a federal exemption, capped at 2,500 vehicles a year, before it could charge fares in its controls-free robotaxi.25 Tesla took a different route and self-certified that the Cybercab meets federal safety standards. NHTSA responded with an Audit Query and then a Special Order requiring sworn answers, now due October 30.2528 Neither step is a finding that Tesla broke the rules, and the Cybercab is still carrying riders in Austin.25 Separately, NHTSA is weighing a petition alleging that about 1,000 Tesla robotaxis sped through school zones. The evidence so far is a few videos, with no crashes reported, and a decision is due around January 21.23

Waymo is not free of scrutiny either. NHTSA told AV developers to submit plans by July 31 to stop robotaxis from interfering with emergency scenes, a directive widely read as aimed at Waymo. No outcome has been made public.25 Waymo, for its part, cites a September analysis covering more than 270 million rider-only miles that found 82% fewer injury crashes than a human-driver benchmark.47

The overall picture is consistent. Waymo is turning a lead in operating experience into denser, more connected service areas. Tesla is scaling fast but still on supervised software, with its consumer promise pushed back again. Regulators are gradually rewriting standards that assumed a human driver. On current evidence, the competition hinges less on who draws the biggest zone and more on who can fill one with cars that need no supervision.