Uber debuts self driving cars in San Francisco today with two operators, one backup for driving and another to collect and analyze data in a passenger seat
With drivers behind the wheels. — Uber is bringing its self-driving technology to San Francisco.
Context & Ripple Effects
Uber is putting its self-driving fleet onto San Francisco streets with a two-person crew in every car — a backup driver behind the wheel and a second operator in the passenger seat collecting and analyzing data. The data seat signals that Uber treats this rollout as much as a mapping and validation exercise as a ridesharing one.
The launch lands in contested territory: within days the California DMV moves against the program, revoking the vehicles' registrations and forcing Uber to halt the pilot and look at deploying elsewhere. Testing later restarts in the city after an Arizona crash had frozen operations, and by mid-2018 Uber restructures the very job created in this debut, laying off roughly 100 backup drivers in favor of smaller numbers of better-trained "mission specialists."
First-order effects
- Every trip runs with two paid onboard staff — a safety driver plus a dedicated data collector — so each mile costs Uber double crew overhead until the data seat proves its value.
- The California DMV becomes the immediate gatekeeper: its registration enforcement halts the SF fleet within weeks, pushing Uber to shop for friendlier states.
Second-order effects
- Uber's testing footprint fragments across state lines — Arizona and Pennsylvania carry the program while California access is litigated and renegotiated — making regulator relationships a core operational input rather than a compliance afterthought.
- The backup-driver role gets consolidated: the 2018 layoffs and "mission specialist" openings show Uber trading headcount for training depth once the data-collection phase matures.
Third-order effects
- If the pattern holds, autonomous-vehicle testing concentrates wherever permits are easiest, and the onboard human job migrates from generic driving backup to specialized fleet oversight — a template for how AV operators staff up and down through regulatory cycles.
The trend: Self-driving programs are becoming roving operations that follow permissive regulators, while the human-in-the-loop job shrinks and specializes as fleets accumulate data.