In the spring of 2026, a safety manager at a mid-size truckload carrier got a question she had never been asked before: "If we book an Aurora autonomous unit through McLeod, does FMCSA (Federal Motor Carrier Safety Administration) apply hours-of-service rules to it?" She knew the HOS (hours of service) rules cold. She had been enforcing them for eleven years. But she realized, standing at her whiteboard, that she genuinely did not know how the rules applied to a truck with no one in the driver's seat, and she had a dispatcher about to book one.

FMCSA: The Regulatory Clock for Every Dispatch Decision

The Federal Motor Carrier Safety Administration is the federal agency responsible for regulating commercial motor vehicles and the carriers that operate them in interstate commerce. FMCSA does not design trucks or write transportation software. What it does is set the legal operating parameters that every dispatcher, fleet manager, and owner-operator must stay inside: how many hours a driver can drive, when they must rest, what equipment inspections are required, how safety performance is scored, and what happens when a carrier's record reflects persistent noncompliance.

Every AI dispatch plan, every autonomous truck booking, and every AI-assisted route optimization that a fleet runs in 2026 still must stay inside the FMCSA framework. The technology changes how decisions are made and how capacity is managed. It does not change the regulatory envelope that decisions must stay inside. This is one of the program's core principles, and this lesson is where that principle gets its technical foundation.

The two most operationally consequential pieces of FMCSA's framework for a dispatcher are the hours-of-service rules and the ELD (electronic logging device) mandate. Understanding them thoroughly is not just a compliance requirement. It is the prerequisite for using AI tools in dispatch without creating liability that the AI optimizer will not warn you about.

Hours of Service: The Rules Every AI Dispatch Plan Must Respect

HOS rules have governed commercial driver working time since the 1930s. They have been updated multiple times, most significantly in 2003, 2011, 2013, 2017, and 2020. Each update was a response to data on driver fatigue as a factor in crashes. The current framework reflects decades of research and litigation, and FMCSA treats violations with corresponding seriousness.

The core rules for property-carrying commercial drivers in interstate commerce are:

The 11-hour driving limit: A driver may not drive a commercial motor vehicle (CMV) for more than 11 hours in a single duty period. Hours spent driving count toward this limit the moment the vehicle is in motion.

The 14-hour on-duty limit: A driver may not drive or perform any on-duty activity after having been on duty for 14 consecutive hours, measured from the time they came on duty following the most recent required off-duty period. The 14-hour clock does not stop for off-duty breaks shorter than the required sleeper berth split or the full restart. A driver who spends 3 hours loading freight at a shipper has already burned 3 hours of that 14-hour window before turning a wheel.

The 30-minute break requirement: Drivers must take a 30-minute break when they have driven for a period of 8 cumulative hours without at least 30 minutes of continuous off-duty time. This break is required within the 14-hour window and does not extend the 14-hour limit.

The 10-hour off-duty requirement: Before a driver can begin a new duty period, they must have at least 10 consecutive hours off duty. This is the reset that restores the 11-hour driving limit and the 14-hour on-duty window.

The 60/70-hour rule: A driver may not drive after accumulating 60 hours of on-duty time in any 7 consecutive days (for carriers operating 5 days per week) or 70 hours in any 8 consecutive days (for carriers operating 7 days per week). This is the weekly cap that catches fatigued drivers who have stayed within daily limits but accumulated dangerous cumulative hours.

The 34-hour restart: A driver who has reached the 60/70-hour limit can restart their weekly calculation by taking 34 consecutive hours off duty. After the restart, the 60/70-hour limit begins fresh.

Each of these rules is a hard constraint on any AI-generated dispatch plan. A plan that proposes a 13-hour driving day is illegal, regardless of how elegantly the optimizer constructed the route. A plan that assigns a load to a driver who has 5 hours of 14-hour window remaining on a 7-hour run has a compliance failure baked in. The AI optimizer does not know it is wrong unless it has been given the current HOS status of every driver and programmed to treat those limits as binding constraints, not soft preferences.

HOS Special Rules Worth Knowing

The core rules apply to the vast majority of property-carrying interstate operations, but several special provisions affect how a dispatcher handles specific situations.

The short-haul exception applies to drivers who operate within a 150 air-mile radius of their work reporting location and return to the same location at the end of each duty tour. These drivers are not required to keep an ELD log, though they must still follow the driving and on-duty limits. A dispatcher running a mix of local and long-haul operations needs to know which drivers qualify for this exception and which do not.

The adverse conditions exception allows a driver to extend their driving time by up to 2 hours when unexpected snow, ice, fog, or road closures make it impossible to safely reach a planned stopping point. This exception requires documentation and cannot be used as a routine extension. An AI dispatch plan that assumes adverse conditions will extend every run is misapplying the rule.

The agricultural exception and construction exemptions apply in specific commodity and operational contexts; a fleet primarily running dry van or reefer freight will rarely encounter them but should know they exist.

Sleeper berth operations, which are common in long-haul trucking, have their own rules. A driver in a truck with a sleeper berth can split the required off-duty time into two periods: one of at least 8 consecutive hours in the sleeper berth, and one of at least 2 consecutive hours (either in the sleeper berth or off duty). The combined time must total at least 10 hours, and both periods must be completed before the driver's next driving period begins. Sleeper splits affect how the 14-hour window is calculated, and an AI tool that does not understand sleeper splits will miscount a team driver's available time.

The ELD Mandate: The Rule That Made HOS Enforcement Automatic

For most of the history of HOS rules, enforcement depended on paper logs. Drivers filled out log books that showed their duty status over the previous 8 days. Roadside inspectors reviewed those logs during inspections, and carrier auditors reviewed them periodically. The system was imperfect: experienced drivers and dispatchers knew how to maintain paper logs that appeared compliant, and enforcement was resource-limited.

The ELD mandate, which reached its final implementation phase in December 2019, changed this fundamentally. An electronic logging device is a piece of hardware connected to the engine control module (ECM) of the commercial vehicle that automatically records driving time whenever the engine is in motion above a low speed threshold. Duty status changes (from off-duty to on-duty, from driving to on-duty not driving) are recorded in real time and transmitted to a telematics provider and, when requested, to enforcement officers during a roadside inspection.

The ELD does not enforce HOS compliance proactively. It does not lock the truck's ignition when a driver reaches the 11-hour driving limit. What it does is create an auditable, tamper-evident record of the driver's actual duty status and driving time that is available to enforcement officers within seconds at a roadside inspection and to FMCSA auditors during a compliance review. Attempting to falsify ELD records is a serious violation with severe penalties, far more serious than the underlying HOS violation the falsification was intended to conceal.

For a fleet that is deploying AI in dispatch, the ELD creates an important dynamic: the AI optimizer's plan and the ELD record of what actually happened are now both visible to regulators. If an AI-generated dispatch plan put a driver on the road who was within their HOS limits according to the plan but the ELD shows differently (because the plan was built on stale data, or missed the loading time at the prior stop, or misunderstood the sleeper split), that discrepancy is not just a bad plan. It is a documented compliance failure tied to a specific dispatch decision.

This is why the AUTHORING KIT principle bears repeating in this context: verify every AI-touched HOS calculation before a driver turns a wheel. The ELD will record what actually happened. The optimizer's plan is not the record; it is a proposal. The dispatcher who commits the plan is responsible for the accuracy of the HOS calculation it embeds.

The FMCSA Driverless Rule: The Regulatory Update in Progress

HOS rules were written for commercial drivers. The definition of "driver" in the Code of Federal Regulations, and throughout the HOS regulatory framework, assumes a human being operating the vehicle. FMCSA is currently in the process of addressing what happens to that framework when there is no human driver.

This update is the regulatory clock the lesson references in the chapter title. Aurora's 250,000-plus driverless miles have been accumulated under a combination of existing regulatory frameworks and FMCSA exemptions that permitted driverless operation in specific conditions. As autonomous commercial operations scale, FMCSA faces the task of developing a comprehensive regulatory framework for vehicles that have no human in the cab. As of mid-2026, that framework is in active development, and fleet professionals need to understand both what is settled and what is still in motion.

What Is Settled

A fully autonomous vehicle (one operating without any human driver or safety driver in the cab) does not have a human driver accumulating HOS fatigue. The HOS rules that cap driving time, require rest breaks, and impose the 60/70-hour weekly limit are fatigue-management tools designed to protect a human nervous system from the consequences of sustained wakefulness while operating heavy machinery. An autonomous system does not fatigue in this physiological sense.

FMCSA has confirmed this principle: HOS rules as currently written do not apply to autonomous vehicle systems in the same way they apply to human drivers. Aurora's driverless trucks do not generate HOS logs for a human driver because there is no human driver to log. The ELD mandate, similarly, was designed to record human driver behavior; an autonomous vehicle's operational records are generated differently and governed under different provisions.

This does not mean autonomous vehicles operate without any regulatory constraints. The constraints are different: they focus on the vehicle's operational design domain, maintenance and software integrity requirements, incident reporting obligations, and the carrier's accountability for ensuring the vehicle operates within its validated parameters. These are not HOS constraints; they are a different regulatory layer that is still being developed.

What Is Still in Motion

FMCSA is working through several open questions in 2026, and a fleet professional operating in this space should monitor these developments through FMCSA's rulemaking notices:

The carrier accountability framework: When a driverless truck is operating on a public highway, the carrier that booked the load (through McLeod TMS, for example) and the company that operates the autonomous vehicle (Aurora) both have some accountability relationship to FMCSA. The precise contours of those accountability relationships, and what obligations the carrier has when it books autonomous capacity, are being defined through FMCSA's rulemaking process. Carriers that integrate autonomous capacity now should stay current on those developments.

Safety reporting requirements: FMCSA requires carriers to report certain incidents involving CMVs. How those incident reporting requirements apply when the CMV involved in an incident is autonomous, and who has the reporting obligation, is part of the regulatory development process.

Inspection and maintenance standards: The DVIR (driver vehicle inspection report) is currently a human driver's document: a record of the pre-trip and post-trip inspection that a CDL holder performs on the vehicle. There is no human driver to complete a DVIR on an autonomous unit. FMCSA is developing the maintenance and inspection standards that apply to autonomous CMVs, which are different from the driver-inspection framework.

Minimum risk condition requirements: When an autonomous system encounters a situation it cannot safely handle, federal guidance requires it to achieve a "minimum risk condition," typically pulling off the road safely. How this requirement interfaces with carrier operations, incident response, and FMCSA reporting is an area where the regulatory details are still being refined.

The lesson's key point on the driverless rule is this: do not assume the regulatory void means no constraints. It means a different set of constraints, currently in development. A carrier that operates autonomous capacity as though FMCSA has no interest in the matter is not correctly reading the regulatory landscape.

How HOS and ELD Apply to AI-Assisted Dispatch

Even setting aside autonomous trucks entirely, AI-assisted dispatch for human-driven fleets creates new ways to both improve HOS compliance and to create new compliance risks. Understanding both sides is essential for any fleet using AI tools in dispatch.

On the improvement side, an AI dispatch optimizer that correctly ingests real-time ELD data for every driver in the fleet can surface HOS status more accurately than any dispatcher tracking a dozen drivers by memory and whiteboard. An optimizer that knows Driver A has 4.5 hours of driving time remaining, Driver B has 9 hours, and Driver C is in their 10-hour reset will propose load assignments that respect those limits. A dispatcher making the same calculation manually, under time pressure, for a fleet of 30 drivers, will occasionally miscalculate or miss an edge case. AI can reduce that error rate significantly, which is a genuine compliance benefit.

On the risk side, an AI optimizer that pulls stale ELD data (from a feed that updated 3 hours ago rather than in real time) will calculate available hours based on where a driver was 3 hours ago, not where they are now. A driver who spent those 3 hours unloading at a customer site has burned 3 hours of their 14-hour window that the optimizer does not know about. The resulting dispatch plan may look HOS-compliant but will put the driver in violation before the delivery is complete.

The verification gate that protects against this risk is the dispatcher's responsibility: before committing a dispatch plan, verify the HOS status in the plan against the current ELD record. If those numbers do not agree, do not dispatch until they do. An AI plan that has not been verified against live ELD data is a proposal, not a compliant dispatch.

This verification requirement also applies when AI tools are generating route modifications on the fly. A real-time reroute that adds 2 hours to a run because of a traffic event may push a driver into HOS violation if the original plan was within the limit by a narrow margin. The dispatcher who accepts the reroute suggestion needs to recalculate the HOS impact before committing the modified route. An AI tool that does this automatically is valuable; one that does not is a liability waiting to surface in an ELD audit.

CSA Scoring and the Compliance Record

FMCSA's Compliance, Safety, Accountability (CSA) program is the scoring system that tracks a carrier's safety and compliance performance over time. CSA (Compliance, Safety, Accountability) scores are built from data collected at roadside inspections, crash reports, and investigation findings. They feed into FMCSA's Safety Measurement System (SMS), which carriers, shippers, and insurers use to evaluate a carrier's compliance posture.

CSA has seven Behavior Analysis and Safety Improvement Categories (BASICs): Unsafe Driving, Hours of Service Compliance, Driver Fitness, Controlled Substances and Alcohol, Vehicle Maintenance, Hazardous Materials Compliance, and Crash Indicator. Each BASIC is scored separately, and carriers whose scores in any BASIC exceed specified thresholds become subject to increased scrutiny, intervention letters, and potentially investigations or audits.

HOS violations discovered at roadside inspections or in ELD records during audits directly affect a carrier's Hours of Service Compliance BASIC score. A fleet that uses AI dispatch tools to improve efficiency but creates HOS violations in the process (through stale data, incorrect calculations, or plans that ignore the 14-hour window) will see those violations accumulate in the CSA database, affecting the carrier's ability to attract quality shippers, renew insurance at favorable rates, and avoid FMCSA intervention.

This connection between AI-assisted dispatch and CSA scoring is not hypothetical. A carrier whose AI optimizer routinely proposes plans with marginal HOS compliance, and whose dispatchers routinely commit those plans without verification, will eventually see roadside inspection findings that reflect those habits. CSA scores reflect patterns over time, and patterns built by an unverified AI tool are patterns the carrier owns.

Key Takeaways

• FMCSA sets the legal operating parameters that every dispatch decision, AI-assisted or not, must stay inside: HOS limits, ELD requirements, CSA scoring, and inspection standards are not optional or soft constraints.
• The core HOS rules for property-carrying interstate drivers are: 11-hour driving limit, 14-hour on-duty window, 30-minute break after 8 cumulative driving hours, 10-hour off-duty reset, and 60/70-hour weekly limit with a 34-hour restart option.
• The 14-hour on-duty window begins when the driver comes on duty, not when they start driving. Loading, fueling, and waiting time all count against it.
• The ELD mandate creates an auditable, tamper-evident record that is visible to enforcement officers at roadside inspections and to FMCSA auditors during reviews, making HOS compliance effectively automatic to detect.
• An AI dispatch plan built on stale ELD data may appear HOS-compliant but create violations in practice. Verifying AI-generated plans against live ELD data before dispatch is not optional; it is the compliance gate.
• HOS rules as currently written apply to human drivers; FMCSA is actively developing a separate regulatory framework for driverless commercial vehicles in 2026, addressing carrier accountability, incident reporting, inspection standards, and minimum risk conditions.
• CSA Compliance, Safety, Accountability scoring tracks HOS violations that accumulate from roadside inspections and audits, connecting AI dispatch practices to long-term carrier compliance ratings that affect shipper relationships and insurance.
• The cardinal rule applies to FMCSA compliance: a dispatch plan that cannot be run legally is a liability, and the dispatcher who commits it owns that liability regardless of whether the plan was AI-generated.