Tesla insists its robotaxi effort is on the rise. The figures it disclosed this week point in nearly the opposite direction.
During yesterday’s earnings call, CEO Elon Musk worked to convince investors of a service that is expanding into new cities, logging more miles and seating more riders in cars without a human minder. His delivery, though, came across as unusually cautious.
“We’re going as fast as humanly possible in scaling Robotaxi,” Musk said, “while trying to ensure that we do not harm anyone at all and ideally do not even run over a pet.”
That closing remark carries more weight than it first appears. The distance between the sales pitch and the actual picture is this: miles driven have essentially plateaued, the fleet looks like it’s contracting, and the safety story remains unresolved regardless of how often the company repeats “zero notable incidents.”
380,000 miles seems impressive until you stack it against a single Waymo week
Ashok Elluswamy, who leads AI at Tesla, said the firm’s driverless robotaxis have covered 380,000 miles across six cities in two states. Musk pegged the growth at 10 percent per week in miles driven.
Here’s the comparison Tesla would prefer you overlook. Waymo, which combines cameras with lidar and radar as redundant sensors, racks up roughly four million miles every week. A full year into its robotaxi program, Tesla’s cumulative total amounts to less than 10 percent of that weekly number.
The gap didn’t escape Wall Street. BNP Paribas Equity Research senior analyst James Picariello described it as “an alarming decline” in a note to clients.
Elluswamy still found room to knock lidar
The sensor debate defines Tesla’s identity, so naturally it surfaced. Elluswamy couldn’t pass up a jab at competitors who rely on lidar rather than Tesla’s camera-only design.
“Historically, the so-called experts have always claimed that you need lidars, radars, HD maps, and the entire kitchen sink to drive safely,” he said. “Here we show that such is not true. You can have safe, comfortable, and affordable autonomy with just cameras. This record should be a huge validation of Tesla’s entire AI approach.”
The catch is that the very record he’s citing is the one displaying flat mileage and a shrinking fleet. Calling that validation is generous.
The fleet you can genuinely tally
Tesla operates a blend of supervised and unsupervised cars, and that balance keeps shifting. According to the crowdsourced Robotaxi Tracker, the present driverless tally sits at 16 vehicles in Austin, four in Dallas and one in Houston.
That adds up to 21 cars underpinning a program Musk portrays as scaling as fast as humanly possible.
A caller raised the obvious point: why scatter a few vehicles across fresh cities rather than saturate Austin, where the service debuted a year ago?
“The reason we have been expanding across different cities, instead of just doubling down on a single city, is that we want to make sure that our stack is a very general one,” Elluswamy said. “It is a general one. We just want to like, you know, both prove to ourselves and to other folks that it is working across a lot of different cities without too much effort per city.”
New Jersey may slam the door shut
Regulation is the second constraint. New Jersey is considering legislation that would essentially outlaw robotaxis lacking multiple sensors — precisely the radar-and-lidar arrangement Tesla declines to adopt.
Morgan Stanley analyst Andrew Percoco brought it up plainly, observing that “there seems to be some evolving regulations at the state level, potentially around sensor requirements.”
Tesla’s vice president of vehicle engineering Lars Morvay made no attempt to spin it. “Obviously the stuff in New Jersey, you know, is a little bit disheartening,” he said.
A patchwork of sensor requirements narrows the market Tesla is able to reach. It also chips away at the narrative of a versatile service capable of driving anywhere in any conditions.
The pet comment was pointed at Waymo
Musk’s quip about sparing someone’s pet wasn’t offhand. It landed as a swipe at Waymo, which has struck and killed several animals over the last few years.
Tesla’s own record is tougher to parse, and that opacity is partly deliberate. The company blacks out many key details in its robotaxi crash reports, leaving outsiders with little to analyze.
What’s available: the Robotaxi Tracker logs about 18 incidents tied to the fleet, with 56 percent of them attributable to another driver. Tesla’s Full Self-Driving is also being investigated by the National Highway Traffic Safety Administration, with a potential recall on the horizon.
One further point worth pausing on. Tesla’s robotaxis are running FSD version 15, while the vehicles sitting in customers’ driveways are on version 14.3.6.
What to track instead of the sales pitch
Tesla is urging investors, fans and customers to overlook all of this and take on faith that the program is expanding. Its own numbers keep pushing back.
To gauge whether the robotaxi narrative holds up, tune out the city tally and the weekly-growth figures. Keep an eye on two signals: whether that 21-car driverless fleet grows in any meaningful way, and whether New Jersey’s bill becomes the model other states adopt. Those shift before the press releases do.