brain/
questionopenautonomous-driving

Does remote-supervised robotaxi qualify as driverless?

Notes

Does remote-supervised robotaxi qualify as driverless?

The question

When a robotaxi has no driver in the vehicle but is monitored (and potentially remote-controlled on intervention) by human operators off-site, should that count as "driverless" / "unsupervised" autonomy? How do regulators, users, and observers draw the line?

Why it matters

Tesla markets its Austin/Dallas/Houston service as "Unsupervised Robotaxi," but the vehicles retain remote supervision 2026-04-20-autoresearch-tesla-fsd. waymo also uses remote assistance — but only on exception, not continuous monitoring (exact division between the two is unclear from available sources). California's CPUC has already drawn a line: Tesla's service is a chauffeur / TCP (limousine) service, not an autonomous vehicle service, regardless of who is in the front seat 2026-04-20-autoresearch-tesla-fsd.

What we currently believe

  • Terminology ("unsupervised," "driverless," "autonomous") is being used inconsistently across operators and regulators.
  • The regulatory category in California (AV permit vs TCP permit) matters more than the marketing label.
  • Remote-supervision operating ratios — how many remote humans per vehicle, how often they intervene — are largely undisclosed by every operator.

Evidence we have

Evidence we need

  • Remote-supervision intervention rates for Tesla and waymo (minutes per ride, interventions per 1000 miles).
  • Remote-operator fleet ratio for each operator.
  • Whether other jurisdictions (Texas, Arizona, Nevada) are drafting their own definitions.

How to resolve

  • Track regulatory filings and disclosures from both Tesla and Waymo.
  • Watch for a state or federal attempt at a standardized "driverless" definition.

Related

Referenced by