Waabi’s Dallas–Houston Autonomous Lane Needs a Human-Handoff Ledger
Waabi and Warrior Logistics have moved autonomous trucking another step out of the demonstration phase and into recurring freight operations. The companies are hauling customer loads between Dallas and Houston for food and beverage, consumer goods, and other shippers, adding autonomous capacity to an established carrier network.
The early results are notable. FreightWaves reports that the operation has achieved 100% on-time pickup and delivery and zero reported safety incidents since starting earlier in 2026. Waabi also says its commercial load volume on the corridor has increased more than 80% this year, although it did not disclose the baseline, fleet size, or number of loads.
Those figures show operational promise, but they do not yet describe the hardest part of autonomous freight: the moments when responsibility moves between a person, a system, a facility, and another person. Before a carrier or shipper scales the lane, it needs a human-handoff ledger that records every transfer of authority and custody.
Map the lane as a chain of controlled handoffs
The Dallas–Houston movement is not one continuous autonomous event. It begins with a shipment plan and ends with accepted freight, with several control points in between. Each point should create a timestamped ledger entry tied to the tractor, trailer, shipment, route, and responsible party.
At the origin yard, the record should confirm trailer identity, seal number, load weight, cargo condition, vehicle inspection, route version, weather review, and dispatch authorization. The person releasing the truck should attest that the physical equipment matches the digital plan. If any field is incomplete or inconsistent, the load should remain blocked.
At highway entry, the ledger should capture the exact moment the autonomous driving system assumes the driving task, its software and map versions, localization confidence, and the status of redundant braking, steering, power, and communications. That event separates yard responsibility from the autonomous segment and makes later analysis far more precise than a generic departure timestamp.
At destination, the sequence reverses. The system should record autonomous disengagement, human acceptance, yard movement, seal verification, proof of delivery, and any damage or shortage. Custody should never become implicit merely because the truck crossed a geofence.
Make intervention evidence useful
Waabi’s current trucks are not yet driverless. The FreightWaves report says the company plans to remove the driver only after the truck and its fully redundant platform are ready, with driverless operations targeted for 2027. That makes today’s supervised service an opportunity to build the evidence model needed for tomorrow.
Every human intervention should be classified, not merely counted. The ledger needs the trigger, location, operating conditions, requested action, actor, response time, duration, outcome, and whether the event affected service or safety. Categories might include construction, emergency vehicles, law-enforcement direction, sensor degradation, uncertain road geometry, mechanical condition, communications loss, or facility access.
Remote support deserves the same discipline. A remote specialist may provide information, approve a revised route, or authorize a minimum-risk maneuver, but those actions are not interchangeable. The record should distinguish advice from control and identify exactly what the autonomous system decided after receiving it.
This level of detail prevents a misleading metric. A lane can report a low intervention count while hiding a small number of lengthy or operationally serious exceptions. Conversely, frequent brief confirmations may indicate workflow friction rather than unsafe driving. Both need different corrective action.
Treat inspections as operating events
Autonomous trucks still have to satisfy roadside and vehicle-safety requirements. FreightWaves has documented the use of Commercial Vehicle Safety Alliance enhanced inspections on the I-45 corridor, designed to verify vehicle condition before an autonomous trip and communicate inspection status electronically to enforcement.
The handoff ledger should therefore connect the inspection result with the dispatched vehicle configuration. It should retain the inspector or authorized technician, inspection standard, defects found, repairs completed, clearance time, and electronic credential status. If a tractor or trailer changes after inspection, the dispatch authorization should expire automatically.
Inspection delay also belongs in lane performance. A safety process may work exactly as intended yet still make a service uneconomic if queues or reinspection rates are excessive. Separating inspection dwell from driving time shows where operational redesign is needed without encouraging teams to weaken the safety gate.
Score the lane with four linked measures
On-time performance matters, but it should sit beside measures that reveal how the service achieved it:
- Intervention rate: human interventions per 1,000 autonomous miles, segmented by cause, severity, and duration.
- On-time performance: pickup and delivery compliance against the customer appointment, with autonomous and human-controlled delays separated.
- Inspection delay: median and 95th-percentile minutes from inspection start to dispatch clearance, plus defect and reinspection rates.
- Cost per autonomous mile: equipment, supervision, remote support, inspections, terminal labor, repositioning, insurance, maintenance, and exception cost divided by productive autonomous miles.
For context, another Texas deployment reported by FreightWaves achieved 100% on-time delivery and about 92% autonomous route coverage in early trials. That second percentage is critical: service reliability alone does not show how much of a trip autonomy actually performed. Waabi and Warrior should pair their on-time result with autonomous coverage and intervention data as the lane scales.
The measures should share a common trip ID. Otherwise, finance may calculate cost on one population of miles while safety and operations report on another. A shared record also lets leaders compare time periods, equipment versions, facilities, customers, weather, and exception types without reconstructing events from separate systems.
Build evidence before removing the driver
The strongest case for autonomous freight will not come from a single perfect percentage. It will come from a repeatable record showing who authorized each movement, what the system encountered, when a human became involved, how custody changed, and what every exception cost.
CXTMS can connect dispatch approvals, inspections, route events, exception workflows, and proof of delivery in one shipment record, giving carriers and shippers an auditable foundation for autonomous operations. Request a CXTMS demo to see how controlled handoffs can support safer, more scalable freight execution.
