GLOBAL DEPLOYMENT TIMELINE · 29 APPROVED TIMELINE EVENTS
Complete approved deployment chronology. Launches, tests, corrections and suspensions use the same evidence ledger; only an explicit stage label changes maturity.LIVE EVIDENCE TRACKER · VERSION 1.3
Autonomous Mobility
Tracking the real-world progression of autonomous vehicles—and what it means for motor insurance through a pricing actuary’s lens.
Regulatory approvals, commercial deployment, fleet size, rides and autonomous mileage are updated as the evidence changes. The Pricing Actuary’s Lens changes only when that evidence materially alters the insurance implications.
01 · GLOBAL MAP
Where is driverless mobility real?
Markers represent the city, service area or regulatory jurisdiction—not an entire country. Stage reflects the most advanced verified public-road deployment in that location.
02 · DEPLOYMENT PROGRESS
How quickly is deployment progressing?
Verified developments are shown separately from stage transitions, so an announcement never looks like operating progress.
TRACKED MARKETS BY STAGE
Counts are records in this tracker, not an estimate of all global programmes.LATEST DISCLOSED SCALE
These measurements come from approved event records. Network totals are not presented as city figures.
RECENTLY PROGRESSED
Waymo lists Atlanta as serving riders
Waymo lists Austin as serving riders
Waymo lists Dallas as serving riders
03 · LATEST DEVELOPMENTS
What changed recently?
Only developments capable of changing deployment, exposure, liability, safety evidence or insurance economics are included.
Singapore stage aligned to verified trial evidence
The official evidence supports regulated testing, but not a current public passenger pilot. Singapore is therefore classified as Stage 1 pending stronger evidence.
Waymo lists Tampa as serving riders
Waymo's current service directory lists Tampa under “Serving Riders In”.
Waymo lists San Diego as serving riders
Waymo's current service directory lists San Diego under “Serving Riders In”.
Waymo lists San Antonio as serving riders
Waymo's current service directory lists San Antonio under “Serving Riders In”.
Waymo lists Orlando as serving riders
Waymo's current service directory lists Orlando under “Serving Riders In”.
Waymo lists Nashville as serving riders
Waymo's current service directory lists Nashville under “Serving Riders In”.
Waymo lists Miami as serving riders
Waymo's current service directory lists Miami under “Serving Riders In”.
Waymo lists Las Vegas as serving riders
Waymo's current service directory lists Las Vegas under “Serving Riders In”.
Waymo lists Houston as serving riders
Waymo's current service directory lists Houston under “Serving Riders In”.
Waymo lists Denver as serving riders
Waymo's current service directory lists Denver under “Serving Riders In”.
Waymo lists Dallas as serving riders
Waymo's current service directory lists Dallas under “Serving Riders In”.
Waymo lists Atlanta as serving riders
Waymo's current service directory lists Atlanta under “Serving Riders In”.
04 · THE PRICING ACTUARY’S LENS
What does this mean for motor insurance?
The framework is unchanged. Each assessment records its direction, confidence, evidence and the future development that would justify a revision.
01Claim frequency
How could this affect claim frequency?
DecreasingMedium confidence+
Claim frequency
How could this affect claim frequency?
A credible reduction in human-error collisions is emerging in constrained operating domains, but it cannot yet be generalised to the whole motor book.
Scaled services now produce meaningful driverless mileage, while the evidence remains concentrated in selected geographies, road types and weather conditions.
Observed mileage is growing, but selection of operating domains and operator-reported comparisons limit transferability.
Independent, exposure-matched loss experience across several cities and adverse conditions.
02Claim severity
How could this affect claim severity?
MixedMedium confidence+
Claim severity
How could this affect claim severity?
Lower collision frequency may be partly offset by expensive sensors, specialist repair, downtime and a larger product-liability component.
AV fleets concentrate high-value hardware and software in intensively utilised vehicles; credible insurer-level severity experience remains limited.
The repair-cost mechanism is clear, but mature claims distributions are not public.
Credible repair invoices, total-loss rates and injury-severity distributions by autonomous mode.
03Exposure basis
Is the exposure unit we currently use still appropriate?
IncreasingHigh confidence+
Exposure basis
Is the exposure unit we currently use still appropriate?
Vehicle-years alone will become progressively less informative. Risk must distinguish human and autonomous miles, operating domain, utilisation and software state.
The same vehicle can move between modes and operating domains with materially different risk controls and responsible entities.
The exposure mismatch follows directly from mixed-mode operation and is visible before full claims credibility emerges.
Standardised, auditable reporting of autonomous miles and system state at policy and incident level.
04Large loss & accumulation
How could this reduce or increase large and accumulated catastrophe losses?
MixedMedium confidence+
Large loss & accumulation
How could this reduce or increase large and accumulated catastrophe losses?
Individual crash risk may fall while correlated technology failures create a new accumulation channel across many vehicles at once.
Common software, mapping, cloud, GNSS, sensor and OTA dependencies can bypass the geographic diversification assumed in conventional motor portfolios.
The dependency structure is observable; the insured loss distribution is not.
A fleet-wide operational incident, systemic recall or insured loss event with common technological cause.
Software defectFaulty OTA updateMapping errorCloud outageGNSS disruptionCyber compromiseSensor defectShared AI modelManufacturer issue
05Price response & selection
How price sensitive is the market to these changes?
UncertainLow confidence+
Price response & selection
How price sensitive is the market to these changes?
Retail customers may expect lower prices for safer automation before insurers can isolate the genuine autonomous exposure benefit.
Consumer pricing remains attached to the vehicle and user while risk ownership is shifting towards fleets, operators and manufacturers.
There is not yet a mature, contestable personal-lines market for fully driverless exposure.
Published purchasing behaviour, tender data or price-elasticity evidence for autonomous mobility cover.
06Capital vulnerability
Could naive insurance capital get burnt in the short to medium term?
IncreasingHigh confidence+
Capital vulnerability
Could naive insurance capital get burnt in the short to medium term?
Yes. Capital can overprice headline safety gains and underprice repair severity, liability ambiguity, sparse data and systemic correlation.
Fast deployment creates pressure to price ahead of credible claims experience, precisely when technology and responsibility are changing together.
The mechanism is consistent with prior new-risk cycles even though ultimate loss ratios are not yet observable.
Stable multi-year loss costs, clearer liability allocation and proven reinsurance treatment.
07Consumer expectation
Do consumers want and expect change?
IncreasingMedium confidence+
Consumer expectation
Do consumers want and expect change?
Usage growth in live markets shows real demand, but acceptance remains sensitive to safety events, availability, trust and price.
Paid ride volumes have scaled materially in live US and Chinese markets, while access elsewhere remains pilot or testing only.
Behaviour is observable in operating markets but is not automatically transferable to the UK.
Sustained repeat usage, willingness-to-pay evidence and post-incident retention across diverse markets.
08Consumer Duty
Knowing this and doing nothing, does it remain consistent with Consumer Duty?
IncreasingMedium confidence+
Consumer Duty
Knowing this and doing nothing, does it remain consistent with Consumer Duty?
Firms should at least identify foreseeable product, pricing and claims-process harms. The required action depends on the exposure actually written.
Mixed-mode vehicles and shifting liability can create unclear cover, poor claims journeys or prices that no longer reflect the service delivered.
The duty is established; the specific response depends on product design and evidence of foreseeable harm.
Meaningful UK customer exposure, unclear policy wording or evidence that current journeys produce poor outcomes.
The assessment applies the existing eight-question framework. Read the framework
05 · EVIDENCE LIBRARY
Why should you believe this?
Material claims should resolve to an original source. Primary regulators and governments take precedence; estimates are labelled rather than presented as fact.
06 · CHANGE LOG
How has the assessment evolved?
Evidence, map and Lens changes are recorded separately. New information is not the same thing as a new actuarial conclusion.
Global tracker
Map, timeline, developments, evidence and headline metrics moved to one approved event ledger. Coverage expanded to 21 visible locations and the full chronology is now shown.
- Map
- Singapore corrected to Stage 1; 12 additional verified US services added
- Lens
- No Lens conclusion changed
Global tracker
Initial evidence-led Autonomous Mobility tracker published with nine verified markets.
- Map
- Initial classification published
- Lens
- Eight-question baseline published
Singapore
Waymo mapping and supervised-test plan added as a candidate future deployment.
- Map
- No stage change
- Lens
- No material change
Global tracker
Timeline, freshness dates and disclosed-scale indicators changed to derive from approved records and cited evidence.
- Map
- No stage change
- Lens
- No material change
Abu Dhabi
Commercial status reconciled to the verified removal of the safety operator.
- Map
- Stage changed from 3 → 4
- Lens
- No material change
Dubai
Commercial status reconciled to the verified fully driverless paid launch.
- Map
- Stage changed from 3 → 4
- Lens
- No material change
Great Britain
Implementation pathway added following the Automated Vehicles Act framework and supervised testing evidence.
- Map
- London Stage 1 baseline classified
- Lens
- Exposure-unit commentary updated