Unassigned Driving Time on ELDs: Review and Annotation Process

A compliance investigator asks for the unidentified driver report covering the last six months. It lists 412 segments, 61 of them longer than 20 miles, and almost every one carries the same note: "not my driving". The safety manager is fairly sure most were shop road tests and yard shunts. Nobody can show it, because the work orders, gate logs and dispatch records were never tied to the segments while the trail was fresh.

A written review process changes that outcome. Each segment is sorted within a fixed number of days, matched to evidence the fleet already keeps, and then either moved to the person who drove it, with that driver's electronic confirmation, or annotated with a reason someone else can check. The working target is plain: stay under the ELD's 30-minute unidentified driving diagnostic and leave no segment without a driver or a verifiable explanation.

• What this guide covers: where unassigned segments come from, the duties in 49 CFR 395.32 and 395.30, a triage query, the assignment handshake, annotation wording, queue ownership and a review checklist.
• What it adds: a timestamped audit trace, a costed fleet example in US dollars with an Indian rupee contrast, and a frank account of why India has no equivalent rule.

Where unidentified driving segments come from in a working fleet

Most unassigned driving is movement the ELD cannot attribute, not hidden hours. Each common source leaves its own trail in another record, so a reviewer who knows the source also knows which document settles it. The list below pairs the usual causes with the evidence that proves or disproves them.

• Shop road tests: a technician drives a tractor after a brake or steering repair without logging in. The repair order, its close time and the technician's badge-in record place the move.
• Yard shunts without status: a driver or yard hand repositions a trailer and never selects the yard move category. The gate log and yard jockey assignment show who held the keys.
• Late or missed login: the driver starts rolling before signing in. The dispatch assignment, a fuel card swipe and the driver's later login at the same spot usually identify the person.
• Display paired to the wrong unit: a tablet moves between tractors, so the driver's log shows one truck while another truck records unidentified miles at the same time.
• Team seat swaps: one co-driver drives while the ELD still shows the other. Section 395.30(c)(3) allows the time to be reassigned between the two if each listed the other as co-driver.
• Device and reconnection artefacts: short segments recorded after a gateway reset or wiring fault. Treat these as a device ticket for the ELD provider and keep the ticket number.

The sources carry different risk. A 0.4-mile shunt inside a fenced yard is a paperwork problem. A 38-mile segment on an interstate at 02:00, on a tractor dispatched to a named driver, may be driving that would have broken an hours limit had it been logged. The review process has to tell those two apart quickly, which is why sorting comes before any edit. Fleet-wide context for this work sits in our logistics and supply chain library.

What 49 CFR 395.32 and the ELD specification require, line by line

The rule gives the driver one duty and the carrier three. The ELD technical specification in Appendix A to Subpart B of Part 395 adds the prompts and the diagnostic that make those duties visible. Reading them together avoids two common mistakes: believing the driver's rejection closes the matter, and believing the carrier may simply clear old records.

Two neighbouring provisions shape every assignment. Section 395.30(d) says edits requested by anyone other than the driver take effect only after the driver confirms them, and a carrier may not request edits before the driver has submitted the records. Section 395.30(f) bars the carrier from altering or erasing the original hours-of-service data or source data streams. Unassigned time can be moved or explained; it cannot be removed.

Unassigned driving review under 49 CFR 395.32(c)(1) leaves a motor carrier two outcomes for every unidentified ELD segment: assign it to the driver, who must confirm under 395.30(d), or annotate why it remains unassigned. Records stay retained for at least 6 months under 395.32(c)(2), and the original data may not be erased.

The ELD technical specification (section 4.3.2.8.2) sets a limit that catches many reviewers out. Unidentified time may be transferred to a driver, and driving time may move between members of a team, but recorded driving time may not otherwise be reassigned and may never be cut. Our guide to locking dispatch against HOS clocks covers how a TMS stops a planner from booking a load the driver cannot legally run. This article deals with what happens afterwards, when the ELD records movement it cannot pin to anyone.

Sorting each segment before anyone edits a log

Sort first, edit second. A reviewer who opens segments in date order and edits as they go will spend the same five minutes on a yard shunt as on an overnight highway run. Putting every segment into one of five bins, using facts the fleet already stores, puts the risky ones at the top of the queue.

1. Bin A, yard only: the whole segment sits inside a geofence the carrier has defined as private yard, under 1 mile and under 15 minutes.
2. Bin B, shop move: the unit had an open repair order that closed within 2 hours of the segment end.
3. Bin C, likely driver: a driver was dispatched on the unit within 3 hours either side, or that driver logged in on the unit within 60 minutes after the segment.
4. Bin D, overlapping log: another driver's log shows driving on a different unit at the same time, which points to a paired display problem.
5. Bin E, unexplained: none of the above; these go to the safety manager the same day.

The query below shows the logic in a form most ELD data exports can feed. Table and column names are illustrative; the thresholds are the ones a carrier should write into its own procedure and review once a quarter.

-- Bin each unidentified driving segment for review (illustrative schema)
SELECT s.segment_id, s.unit_id, s.start_utc, s.end_utc, s.miles,
  CASE
    WHEN s.inside_private_yard = 1 AND s.miles < 1.0
         AND s.minutes < 15                              THEN 'A_YARD'
    WHEN EXISTS (SELECT 1 FROM repair_orders r
                 WHERE r.unit_id = s.unit_id
                   AND r.closed_utc BETWEEN s.end_utc - INTERVAL '2 hours'
                                        AND s.end_utc + INTERVAL '2 hours')
                                                          THEN 'B_SHOP'
    WHEN EXISTS (SELECT 1 FROM dispatch_assignments d
                 WHERE d.unit_id = s.unit_id
                   AND s.start_utc BETWEEN d.planned_start_utc - INTERVAL '3 hours'
                                       AND d.planned_end_utc + INTERVAL '3 hours')
                                                          THEN 'C_DRIVER'
    WHEN EXISTS (SELECT 1 FROM eld_driving_events e
                 WHERE e.unit_id <> s.unit_id
                   AND e.start_utc < s.end_utc AND e.end_utc > s.start_utc
                   AND e.display_id = s.last_display_id)    THEN 'D_PAIRING'
    ELSE 'E_UNEXPLAINED'
  END AS review_bin
FROM unidentified_segments s
WHERE s.review_status = 'OPEN'
ORDER BY s.miles DESC;

Order by miles, not by date. A queue sorted by distance surfaces the segments that could hide an hours violation, and those are the ones an investigator samples first. Bin A and Bin B entries still need an annotation each, but they can be worked in batches once the evidence is attached.

Moving a segment to a driver: the confirmation handshake

Assignment is a proposal until the driver accepts it. Once accepted, the driver's clocks are recalculated, and any violation the extra driving creates stands on the record. Leaving a segment unassigned to avoid that result is exactly what an audit is designed to find.

• Carrier proposes: the safety analyst selects the segment, picks the driver and adds a note citing the evidence, such as dispatch trip number or fuel transaction.
• Driver decides: at next login the ELD shows the proposed change, and the driver confirms or rejects it, then recertifies the affected day.
• Clocks recalculate: the added driving counts toward the 11-hour, 14-hour and weekly limits for that day, which may flag a violation after the fact.
• Rejection is not closure: if the driver rejects and the evidence is strong, the carrier records its findings in an annotation and treats the matter as a driver performance or training issue.

Section 395.30(e) prohibits coercing a driver into falsely certifying records, so the conversation after a rejection has to be about evidence, not pressure. Keep the call notes. If the evidence is weak, the honest annotation says so.

The trace below follows one segment through an uncontrolled week and then shows what a working process would have done. Unit numbers, names and times are illustrative.

1. Mon 20:50 CT, driver log: driver R. Ortiz, on duty since 07:05, goes off duty on unit 2318 at a Fort Worth customer without selecting personal conveyance and logs out.
2. Mon 21:05 CT, unit 2318: the tractor drives 38.4 miles to the terminal with nobody logged in; the ELD writes segment UDS-55120, 53 minutes.
3. Tue 05:40 CT, login prompt: Ortiz logs in, sees the unassigned warning and rejects it. The segment stays under the unidentified profile with no note.
4. Thu 14:10 CT, weekly review: the analyst finds dispatch trip 7781 still open on unit 2318 at 21:05 and the terminal gate scan at 21:58 under Ortiz's badge. Assignment is proposed; Ortiz accepts.
5. Thu 14:12 CT, recalculation: the 14-hour window is now exceeded by 53 minutes on Monday. The violation is logged, coaching is recorded, and dispatch adds a rule blocking terminal returns that would exceed the window.

Driver assignment of unidentified ELD time works as a handshake under 49 CFR 395.30(d): the motor carrier proposes, the driver confirms or rejects, and the day is recertified. A 53-minute segment accepted after review can turn a compliant 13-hour day into a 14-hour window breach, and that breach must stay on the record.

Annotations an investigator can verify months later

A good annotation names the cause, the person if known and the document that proves it. Free text such as "unknown" or "not driver" tells an investigator nothing and invites the assumption that the time hid a violation. Write each note so that someone with no memory of the event can check it against a record in under five minutes.

The last row matters most. The rule does not require a carrier to solve every segment; 395.32(c)(1) expressly allows a note explaining why time stays unassigned. What it requires is a reviewed record. A note that lists the sources searched shows the review happened; a blank or generic note suggests it did not.

Keep annotation templates short and make the evidence reference mandatory. A template that forces a repair order number, a gate row or a ticket number produces notes that age well, and it stops reviewers from typing the same vague phrase into hundreds of events.

Queue ownership, review cadence and the 30-minute diagnostic

One named owner should hold the unassigned queue, with a daily triage and a weekly sign-off. Without an owner, segments age past the point where gate and fuel records are easy to pull, and the per-unit diagnostic stays lit for days. The table sets out who keeps each record and who approves the result.

The diagnostic sets the pace. Under Appendix A section 4.6.1.6(c), more than 30 minutes of unidentified driving in a 24-hour period sets the unidentified driving records diagnostic for the current period and the following 7 days. Section 4.6.1.6(d) clears it only when the total for the current 24 hours and previous 7 consecutive days falls to 15 minutes or less. A single long segment can keep a unit flagged for more than a week.

That arithmetic argues for daily triage of Bin C and Bin E, because assigning time to a driver removes it from the unidentified total. Bins A and B still need annotation, but a weekly batch is usually enough. A roadside inspector who sees the diagnostic indicator on the display will ask about it, so drivers should know what it means and who to call. Other diagnostics and full malfunctions follow different repair and paper-log rules, set out in our ELD malfunction guide.

Pricing the exposure: a worked fleet example in dollars and rupees

Measure unassigned driving as a share of ELD miles, then cost the review. The ratio, unassigned miles divided by total ELD-recorded miles for the same units and period, shows whether the problem is shrinking; the cost shows what the process consumes against what it protects. Figures below are illustrative.

The US column shows a process that costs about $1,200 a month in analyst time and corrects driver pay of a similar size. The larger value is harder to price: a file an investigator can sample without finding blank or implausible notes. For the Indian fleet, the regulatory driver is absent, and the value sits in fuel and trip settlement, which our fleet fuel reconciliation guide covers in depth.

Unassigned driving exposure is best tracked as unassigned ELD miles divided by total ELD miles per month. A 140-tractor fleet at 9,880 of 1,520,000 miles, a 0.65% ratio, spends roughly 28 analyst hours monthly clearing the 49 CFR 395.32 queue, about $1,189 at a $42 loaded rate.

India has no ELD rule, so the same review serves different ends

We found no Indian requirement for electronic logging devices or an unidentified driver profile, so there is no regulatory counterpart to 395.32. Indian fleets that run GPS telematics still see trips with no driver attached, and reviewing them protects money rather than an audit file.

The legal frame for driver hours is different. Section 91 of the Motor Vehicles Act, 1988 ties the hours of work of transport vehicle drivers to the Motor Transport Workers Act, 1961. The Government made the Occupational Safety, Health and Working Conditions Code, 2020 effective on 21 November 2025, consolidating central labour laws including that Act, with existing provisions continuing during the transition. None of this creates a device-level record of who drove.

• Trip sheet matching: each GPS trip should match a trip sheet or lorry receipt with a named driver; unmatched trips go to the transport manager the next day.
• Allowance and incentive pay: per-kilometre allowances paid on trip sheets drift when trips are unattributed; our driver trip settlement workflow shows how to close them.
• Diesel accountability: unattributed kilometres mean unattributed fuel, often the largest single leak in an owner-operated fleet.
• Hours evidence: where state rules under the labour codes require hours registers, GPS trips give supporting evidence, but the register itself remains the legal record.

Which review steps a rule can settle and which need a person

Software can sort, match and draft; a person must decide who drove and sign the note. The split below keeps automation on the tasks where a wrong answer is cheap and visible, and keeps people on the tasks where a wrong answer becomes a false record.

Do not let any tool write annotations straight into the ELD without review. A model that reads "road test" on one repair order and generalises it to similar segments will produce exactly the uniform, implausible notes that an investigator looks for. Drafting saves time; saving the note is a person's act.

Checklist for setting up or auditing the review process

A review process is ready when every open segment has an owner, an age limit and an evidence path. Use the checklist to set one up, or to test an existing one before an audit or a change of ELD provider.

Run the checklist again whenever the fleet changes ELD provider, adds a terminal or reorganises the shop, because each of those changes breaks one of the evidence paths. A sample audit of 20 closed segments each quarter, chosen by highest miles, is a cheap way to see whether notes still match the evidence.

If you are comparing ELD, TMS or maintenance systems and want the unassigned review built into the workflow rather than run from spreadsheets, you can send us your requirements and we will map the data each step needs.