brianletort.ai
← Library

Waymo

Verified evidenceExceptions onlyCreator to judge

Fully autonomous Waymo One driving

Within a defined domain, automated driving can control the vehicle while remote assistance handles exceptions.

Engineering design · Phoenix, San Francisco, Los Angeles, and Austin, United States

Collections: Human still decides · Embodied work

An editorial scene for Waymo contrasts perceives the road, plans maneuvers, and controls the vehicle for each passenger trip. with perceives, predicts, plans, and controls the vehicle with no human in the driver seat. in the fully autonomous waymo one driving workflow.

Executive brief

The operating-model shift, in one view.

Physical autonomy is defensible only when the decision boundary, exception path, and real-world harm metric are explicit.

AI value · Property-damage liability claims

Verified

9 Waymo claims, an 88% reduction.

Claims are a proxy for at-fault collisions and results apply to Waymo's operating domains.

Before

Perceives the road, plans maneuvers, and controls the vehicle for each passenger trip. → Dispatches and monitors a human-driven trip.

After

Perceives, predicts, plans, and controls the vehicle with no human in the driver seat. → Monitors service and provides remote assistance for unusual situations without continuously driving.

Human boundary

The ADS controls routine trips; operations staff and authorities define service boundaries, suspend operations, and manage exceptions.

Why it matters

Within a defined domain, automated driving can control the vehicle while remote assistance handles exceptions.

How the work changed

Before

How the work ran before the change.

  1. Step 1 of 2

    Human driver

    Perceives the road, plans maneuvers, and controls the vehicle for each passenger trip.

    ControlTraffic law, licensing, and driver judgment.

  2. Step 2 of 2

    Ride-hailing operator

    Dispatches and monitors a human-driven trip.

    ControlFleet and safety procedures.

What changed

Within a defined domain, automated driving can control the vehicle while remote assistance handles exceptions.

Decision rightAI handles the default; humans own exceptions

After

How the same work runs now.

  1. Step 1 of 2

    Waymo Driver

    Perceives, predicts, plans, and controls the vehicle with no human in the driver seat.

    ControlGeofenced operational domain and safety systems.

  2. Step 2 of 2

    Fleet operations

    Monitors service and provides remote assistance for unusual situations without continuously driving.

    ControlOperational escalation and regulatory reporting.

Process model built from the published workflow evidence for Waymo. Every step, actor, and control appears in full below.
Every step, actor, and control

Exception path

The vehicle reaches a minimal-risk condition or requests fleet assistance; field teams and emergency services respond when needed.

Work removed

  • Continuous in-vehicle human driving

Decision authority

The ADS controls routine trips; operations staff and authorities define service boundaries, suspend operations, and manage exceptions.

Before

  1. 01

    Human driver

    Perceives the road, plans maneuvers, and controls the vehicle for each passenger trip.

    Control: Traffic law, licensing, and driver judgment.

  2. 02

    Ride-hailing operator

    Dispatches and monitors a human-driven trip.

    Control: Fleet and safety procedures.

After

  1. 01

    Waymo Driver

    Perceives, predicts, plans, and controls the vehicle with no human in the driver seat.

    Control: Geofenced operational domain and safety systems.

  2. 02

    Fleet operations

    Monitors service and provides remote assistance for unusual situations without continuously driving.

    Control: Operational escalation and regulatory reporting.

Work that left the path

  • Continuous in-vehicle human driving

Human role before

A driver made all moment-to-moment driving decisions.

Human role after

Humans design the operating domain, monitor fleet health, assist exceptions, and manage incidents rather than drive each trip.

AI roleDecision mode: bounded autonomy. The ADS holds real-time vehicle-control decisions inside approved service areas.

Outcomes

Property-damage liability claims

Verified

78 claims expected for human drivers over comparable 25.3 million miles.9 Waymo claims, an 88% reduction.

25.3 million fully autonomous miles analyzed through 2024. · Four U.S. cities; human baseline used over 500,000 claims and 200 billion miles.

Claims are a proxy for at-fault collisions and results apply to Waymo's operating domains.

Bodily-injury liability claims

Verified

26 claims expected for human drivers over comparable mileage.2 Waymo claims, a 92% reduction.

25.3 million fully autonomous miles analyzed through 2024. · Four U.S. cities.

Both injury claims were still open at publication.

What leaders can reuse

Anti-pattern

Generalizing geofenced service results to unrestricted autonomous driving.

Questions

  1. 01What conditions are outside the operating domain?
  2. 02How is minimum-risk behavior validated?
  3. 03Which incident data are independently reviewable?

Portability conditions

  • Defined operational domain
  • Redundant safety systems
  • Remote and field response
  • Regulatory reporting

Reputation risk

medium

Evidence and authority

What the public record supports.

Current · updated

1 peer reviewed, 1 vendor; publication outcomes are verified.

Bundle 1.0.0 · reviewed 2026-08-23 · stable ID 0809d69805cefe21

Related transformations

More in Engineering design

Sources

Read the evidence, freshness, caveat, and version policy.