Innov8ion.AI
AI in Fleet Management
Prepared November 9, 2026
AI in Fleet Management Briefing · November 9, 2026

Fleet AI’s Next Test Is Service Quality, Not Task Volume

Two qualifying stories describe different operating decisions: Forterra and May Mobility are advancing to prototype autonomous on-base transportation with a live fleet command view, while Load One has deployed AI agents into expedited freight quoting, dispatch, tracking and billing workflows. The first remains a test program; the second is an operating deployment with company-reported task and processing figures.

Deployment maturity: Forterra and May Mobility advance to prototype testing; Load One reports an operating deployment with company-reported task volumes.

Keep people accountable: The base prototype includes safety operators, remote assistance and safe-stop behavior; Load One routes exceptions and checks to staff.

Measure service outcomes: Compare use, wait times and interventions for shuttles; audit quote fit, margin, overrides and exception quality for freight agents.

Freight operations dispatcher reviews AI-assisted workflow exceptions at a loading dock
Human Review at the AI Control Tower

Executive Readouts

Decision-oriented takeaways from today’s fleet-management scan.

  • Prototype.
  • The Fort Benning pilot should measure vehicle use, wait time, trips completed, personnel hours returned to training, interventions and data controls.
  • Load One says seven agents handle about 3,800 daily tasks; validate task quality, exceptions and staff overrides as volume grows.
  • Its quote agent reportedly returns a rate in about two seconds versus a former five-to-20-minute manual range; test equipment fit, accuracy, acceptance and margin.
  • Audit the dispatch phone agent’s first-week 1,381 confirmations and the billing agent’s reported 275 invoices per 15 minutes against human-checked samples.
  • Keep safety operators, remote assistance and safe-stop controls during autonomous shuttle trials; retain staff review for low-confidence quotes, equipment rules and billing exceptions.

Executive Summary

Two qualifying stories describe different operating decisions: Forterra and May Mobility are advancing to prototype autonomous on-base transportation with a live fleet command view, while Load One has deployed AI agents into expedited freight quoting, dispatch, tracking and billing workflows. The first remains a test program; the second is an operating deployment with company-reported task and processing figures.

General AI in Fleet Management

No additional cross-functional fleet AI development met the inclusion standard for this edition.

Fleet Strategy & Demand Planning

No qualifying new fleet capacity or investment decision was verified for this edition.

Vehicle & Asset Acquisition and Onboarding

No qualifying AI-supported fleet acquisition or asset-onboarding decision was verified for this edition.

Driver & Workforce Readiness

No qualifying AI-enabled driver or technician readiness development was verified for this edition.

Dispatch, Routing & Daily Operations

01Fleet signal

Forterra and May Mobility selected for military-base autonomous shuttle prototyping

Forterra and May Mobility were selected with four other vendors to advance to Phase III of the Department of War Innovation Unit’s Robotic Operations for Autonomous Delivery and Sustainment (ROADS) challenge. The up-to-$30 million program is testing commercial self-driving vehicles for base shuttles and logistics trucks, with goals that include using government vehicles more effectively and reducing the time personnel spend driving routine trips.

May Mobility will supply its autonomous driving system and passenger vehicles; if adopted, riders could request trips through a rideshare app. Forterra will operate the prototype through OneLink, which gives installation leaders a live vehicle map and dashboard with installation-specific parameters. May Mobility describes its onboard model as predicting up to 10 seconds ahead and evaluating multiple driving strategies.

Vendor mapping is planned for November, followed by phased demonstrations through 2026, including a month-long pilot at Fort Benning. The design includes onboard safety operators, remote assistance and conservative safe-stop behavior. Selection advances the teams to prototyping; broader adoption remains contingent on evaluation of technical performance, operating suitability, user needs, safety and data management.

Why it matters:

Bases can have government vehicles sitting idle while personnel are diverted from training to shuttle people and equipment. This prototype tests whether pooled, on-demand autonomous service can make those assets more useful without shifting transport demand onto additional staffed trips.

Practical AI use case or operational implication:

Installation transport managers can use the OneLink map and trip demand to assign vehicles to recurring clinic, range and facility movements, while setting local operating limits and routing exceptions. The planned pilot should retain safety operators and remote assistance while the installation measures service reliability and intervention rates.

Suggested executive takeaway:

Before funding a broader rollout, require the Fort Benning pilot to report vehicle use, passenger wait time, personnel hours returned to training, completed trips, safety interventions and data-handling controls against a current-service baseline.

How fleet operators could use this:

Campus, base and large-site transport teams can first map recurring short trips and current vehicle idle periods, then test a request-based shuttle on a bounded service area. Keep dispatch ownership, rider escalation and safe-stop response with named staff until operational evidence supports reducing those controls.

#AutonomousVehicles #FleetOperations #MilitaryMobility #Dispatch
02Fleet signal

Load One deploys an AI control tower across expedited freight workflows

Taylor, Michigan-based expedited carrier Load One has deployed seven in-house AI agents through a centralized AI Control Tower. The agents support quoting, billing compliance, load tracking and driver dispatch; the company says they handle about 3,800 tasks each day that were previously performed by people.

For a rate request, a language model reads the customer email and passes extracted details to Load One’s Transcend transportation management system. The system checks available company capacity, partner-carrier bids and market density before returning a quote in about two seconds; Load One’s president compared that with a former manual range of five to 20 minutes. Supervisors use human review and predefined rules for equipment details such as dock-height requirements.

Other agents contact drivers about offers, monitor shipment messages and GPS updates, and check delivery paperwork against billing requirements. The company reported 1,381 dispatches handled by a newly deployed load-confirmation phone agent in its first week and said its billing agent can process roughly 275 invoices in 15 minutes. These are company-reported operating figures, and the control tower is designed to route exceptions and quality checks to staff.

Why it matters:

Expedited carriers compete on how quickly they can commit suitable capacity to a time-sensitive shipment. Faster quote preparation can help staff handle more requests, but incorrect equipment matching or a missed exception could turn a speed gain into a service failure.

Practical AI use case or operational implication:

A carrier operations team can connect inbound rate requests to TMS capacity, partner bids and market-density checks, then let staff review low-confidence equipment fits before a quote is committed. Dispatch managers can separately audit driver-call outcomes, GPS exceptions and proof-of-delivery discrepancies.

Suggested executive takeaway:

Load One’s operations leadership should compare AI-generated quotes with historical manual decisions for equipment fit, quote accuracy, acceptance and margin, while monitoring human overrides and exceptions as agent volume grows.

How fleet operators could use this:

Carriers with a mature TMS and frequent expedited quoting can test one email-to-quote workflow where capacity and equipment rules are explicit. Smaller or less-integrated fleets can begin with a narrow call-confirmation or document-check task using existing systems, rather than building a custom control tower before data and exception ownership are ready.

#FleetAI #ExpeditedFreight #Dispatch #TransportationManagement

Safety, Compliance & Incident Management

No qualifying new fleet AI safety, compliance or incident decision was verified for this edition.

Maintenance, Fuel, Parts & Downtime Management

No qualifying new AI-supported maintenance, fuel, parts or downtime workflow was verified for this edition.

Performance, Cost & Sustainability Optimization

No qualifying fleet AI performance, cost or sustainability outcome was verified for this edition.

Replacement, Disposal & Lifecycle Renewal

No qualifying AI-supported asset replacement or lifecycle renewal decision was verified for this edition.

Bottom Line

The base-shuttle program is still in prototype evaluation, while Load One’s workflow automation is already operating at reported daily volume. Executives evaluating either model should tie expansion to measured service, safety and exception-handling results rather than task counts alone.