MARKET CATALYST · MEASUREMENT STARTING POINT
Driverless Transport
Driverless Transport: Can Driver-Out Service Become Repeatable, Safe and Economic?
DATED BASELINE
Starting record
October 22, 2026
Driverless Transport: Can Driver-Out Service Become Repeatable, Safe and Economic?
Waymo reports 271.3 million rider-only miles through June 2026. California deployment records show about 4.22 million completed trips in Q2. Kodiak and Atlas report paid driverless work in the Permian, while Aurora has started driverless highway freight. These are operating records, not proof of fleet-wide profitability or unrestricted autonomy. Supervised demonstrations, factory capacity and future vehicle orders remain separate.
Next review: At the dated evidence gates in the thesis · Terminal review: No fixed terminal date
The baseline records what was known at initiation. It does not assign a realized outcome or grade.
Retain operator, platform, operating domain, jurisdiction and quarter. Separate passenger and freight branches. All six domains have unresolved economics; historical safety results do not automatically transfer to new domains.
UNDER OBSERVATION
U.S. surface-street ride-hailing
- Starting evidence
- Waymo reports 271.3M rider-only miles through June. IIHS supports a historical, bounded safety advantage; Tesla and Zoox use different measures.
- What would move the thesis
- Reconcile CPUC completed trips, occupied/all miles and service availability with matched-domain crash exposure. Track Waymo, Tesla and Zoox separately; do not treat fared flags as collected revenue.
- Next checkpoint
- Next CPUC quarter and operator safety releases; dates unconfirmed
Source record ↗UNDER OBSERVATION
Freeways and airports
- Starting evidence
- New domains are outside the established surface-street safety benchmark.
- What would move the thesis
- Require domain-specific exposure, injuries, service restrictions and interventions before extending a safety conclusion. Record airport authorizations separately from usable paid service.
- Next checkpoint
- At domain-specific operating disclosures
Source record ↗UNDER OBSERVATION
China urban ride-hailing
- Starting evidence
- Apollo Go, Pony and WeRide disclose distinct ride, kilometre, fleet and revenue definitions.
- What would move the thesis
- Keep operator and city records separate; identify paid versus promotional service, same-period fleet exposure and support costs. Do not combine China kilometres with U.S. safety benchmarks.
- Next checkpoint
- Quarterly company disclosures
Source record ↗UNDER OBSERVATION
Permian sand hauling
- Starting evidence
- Kodiak/Atlas reports paid driverless hours and loads; Aurora/Detmar is a separate branch.
- What would move the thesis
- Track active trucks, paid hours, loads, tons and remote/recovery support by operator and period. Verify expansion targets against actual fleet use and domain contribution.
- Next checkpoint
- Quarterly operations; Atlas fleet target mid-2027 remains a target
Source record ↗UNDER OBSERVATION
Middle-mile and regional freight
- Starting evidence
- Gatik and Pony Mawan operate different routes and jurisdictions; contract headlines are not recognized driverless revenue.
- What would move the thesis
- Verify removal of safety operators, route-specific paid operations, active fleet hours and customer continuity. Separate contracted amounts from realized revenue and cost.
- Next checkpoint
- At dated route and financial disclosures
Source record ↗UNDER OBSERVATION
Open-highway Class 8 freight
- Starting evidence
- Aurora has driverless miles but mixed supervised revenue; Kodiak long-haul and Volvo scale-up include future targets.
- What would move the thesis
- Separate driverless from supervised miles/revenue. Seek loaded and empty miles, full support costs, domain contribution and funded replacement economics. Verify year-end targets without assuming completion.
- Next checkpoint
- Q3 filings and year-end 2026 milestones; exact report dates unconfirmed
Source record ↗UNDER OBSERVATION
Safety exceptions and remote assistance
- Starting evidence
- Post-remedy school-bus incidents remain under investigation; intervention frequency is not fully disclosed.
- What would move the thesis
- Follow NTSB findings and recall scope, software version, remote-advice versus control, and response outcomes. Preserve preliminary status and avoid inferring fault or defect recurrence.
- Next checkpoint
- Investigation publication is unscheduled
Source record ↗These are monitoring questions, not completed outcomes. Company targets retain that label; unscheduled disclosures have no invented deadline. Later observations must carry their own date, source and measurement scope.
Record linkagePublication 1000 · Thematic tracker 6. Later observations and judgments remain separate from this starting record.