Intermodal Drayage Chassis Utilization Data: A Research Method

Drayage equipment can be physically present but unavailable because it is at a terminal, assigned to a load, awaiting repair, or missing a usable interchange record. A useful dashboard makes those states explicit.

Methodology

Assign a stable chassis identifier and record gate-in, gate-out, interchange, repair-open, repair-close, and dispatch events. Define available time and exclude periods outside the operator's control only with a documented rule. Reconcile terminal records to dispatch and maintenance systems.

Findings

Measure	Meaning
Turns	completed moves per chassis period
Utilization	assigned or moving time divided by available time
Out-of-service share	repair or defect time divided by observed time

FMC and FHWA materials provide policy and network context, but they do not establish one universal utilization denominator. The method must state whether idle terminal time is available, constrained, or excluded.

Operational interpretation

Publish a daily exception list for missing interchanges, overdue repairs, and equipment with no current location. Keep safety release decisions with the qualified maintenance owner.

Limitations

Intermodal systems use different identifiers and event clocks. A terminal scan may be delayed or duplicated. Small chassis pools create volatile percentages.

Data design

The unit is one container move paired with one chassis assignment. The population should include all planned, completed, canceled, and rolled moves in a named port or rail-served geography during a stated period. Count moves, not chassis-days, when measuring move completion. Count chassis-days separately when measuring utilization. Keep empty repositioning, out-of-service status, maintenance hold, interchange delay, and unavailable chassis as different states. A single “used” flag cannot distinguish demand from supply.

For every move, preserve container number, equipment type, chassis identifier, interchange location, appointment version, ingate and outgate timestamps, pickup and return timestamps, power-unit identifier, carrier, and source system. Normalize local time to an analysis timezone while retaining the original value and timezone. If an event is estimated from a GPS ping, mark it estimated. GPS presence at a terminal is not proof of interchange, inspection, or release.

Evidence and measures

BTS freight statistics, FHWA freight-performance resources, and port or rail-published operating notices provide context for mode flows and network conditions. They do not establish a carrier's chassis supply, terminal queue, or contractual responsibility. Calculate utilization as occupied chassis-days divided by available chassis-days only after defining availability. Report fleet count, occupied days, unavailable days, missing events, median turn time, and the 90th percentile. Do not compare a terminal with a different appointment policy without restating the denominator.

Separate the time from container release to pickup, pickup to ingate, terminal dwell, outgate to delivery, and delivery to return. A long total turn may be caused by one segment while the others remain normal. Analyze by terminal, equipment class, weekday, appointment window, and carrier only where sample size supports comparison. Retain cancellations and reappointments rather than deleting them; they reveal planning friction but should not be mixed into completed-move turn time.

Interpretation

Low utilization can mean surplus chassis, weak demand, or a data gap. High utilization can mean efficient circulation or a shortage that leaves containers unserved. A rising turn time can reflect terminal congestion, appointment scarcity, road conditions, inspection, or a changed event-capture rule. Review reason codes and missingness before assigning cause. Public freight indicators are corroborating context, not a substitute for the move ledger.

The method transfers to drayage fleets, brokers, and intermodal operators when equipment identifiers are durable. It is weaker when chassis are pooled without a reliable interchange record or when terminals overwrite appointment history. The bounded conclusion is that chassis utilization is a state-and-time measurement: it becomes defensible only when move events, availability states, and geography are defined separately. It cannot by itself prove that a terminal, carrier, rail provider, or shipper caused a delay.

Worked interpretation

Consider two terminals with the same occupied chassis-days. One may have high availability and moderate demand; the other may have low availability and repeated missed moves. The utilization percentage alone cannot distinguish them. Pair utilization with unserved move requests, average turn time, empty repositioning, and unavailable-day reasons. Also report the number of physical chassis and the number of distinct identifiers observed so pooled equipment is not mistaken for a dedicated fleet.

Event quality deserves its own score. Count moves with complete interchange, pickup, delivery, and return evidence, then report the complete-record share by terminal and carrier. If completeness falls during a system migration, mark the break and avoid a false before-and-after conclusion. The safest conclusion is often that utilization is measurable for the complete subset while total network availability remains uncertain.

Additional limitation

Pool rules can make one physical chassis appear under multiple operational owners. Reconcile interchange and pool assignment before attributing utilization to a carrier. Where the identity cannot be resolved, publish a pooled-network result and state that carrier-level attribution is unavailable.

Reporting standard

State whether availability is physical, contractual, or system-visible. Report occupied days, available days, unavailable days, unserved requests, and complete-record share together. This prevents a high ratio caused by a shrinking denominator from being described as improved circulation and makes pooled equipment limits visible.

Reproducibility note

Retain the equipment-state history, interchange evidence, terminal scope, and event-completeness flag. Recalculate utilization after pool assignments are corrected, but preserve the prior result as a revision. This is necessary because a changing equipment identifier can otherwise look like a change in circulation.

Separating availability from utilization

Availability is a stock observation; utilization is an exposure measure. A chassis can be available at noon and unavailable by evening, so a daily count cannot substitute for chassis-hours or chassis-days. Define whether the denominator is the pool, the terminal inventory, the number of moves, or the time observed. Report the observation window and missing state transitions. If equipment leaves the data system, do not assume it was productive or idle; mark the interval unknown.

Port comparisons also require comparable boundaries. One terminal may record interchange at the gate, another at the appointment, and a pool may record a maintenance event separately. Align event definitions before comparing percentages. Weather, vessel bunching, customs holds, and street-turn programs can change the mix of moves. These factors may explain a pattern, but they should be described as candidate explanations unless matched records support them.

A useful output includes the population, state definitions, exposure denominator, missingness, and uncertainty around small groups. The evidence supports a measured description of equipment movement and time in state. It does not establish a universal chassis shortage or assign responsibility for a failed appointment. Transfer is bounded to settings with similar data capture and operating rules.

Reporting uncertainty

Publish the pool definition, terminal boundary, observation window, state definitions, count of moves, chassis-days, and missing transitions. A reader should know whether a percentage describes a snapshot or time at risk. Keep unavailable equipment separate from unobserved equipment. This prevents a data gap from being interpreted as productive use or idle capacity and makes the conclusion transferable only to comparable operations.

Bounded conclusion

The evidence supports a terminal-specific description of equipment state and exposure time when identifiers and transitions are complete. It does not establish systemwide capacity or assign responsibility for delay without matched records.

Sources

1. FMC Intermodal Transportation
2. FMC Fact Finding 29
3. BTS Freight Facts
4. BTS FAF
5. FHWA Freight Management
6. FHWA Freight Measures
7. FMCSA Maintenance
8. FMCSA SAFER
9. Census CFS
10. EIA Diesel

Unit of analysis

Chassis utilization is often overstated because the available pool is not defined. A defensible observation identifies the terminal, equipment owner or pool, chassis identifier when permitted, appointment or interchange event, status, and time window. Distinguish a chassis physically unavailable from one reserved, under repair, held by a driver, or present but incompatible with the move. A daily snapshot can describe availability at a moment; an event history can describe how long equipment remained in each state. Those measures should not be substituted for one another.

For a bounded study, use completed drayage moves within named terminals and dates, then report moves, chassis-days observed, missing interchange events, and the share of records with a complete start and end. Segment by terminal, pool, equipment type, and import or export direction where sample size permits. A high utilization percentage may coexist with failed appointments if the denominator excludes unavailable or unrecorded units. Preserve cancellations and no-shows as separate statuses rather than treating them as completed moves.

The evidence can establish recorded equipment state and elapsed time for the selected population. It cannot prove that a terminal caused a delay when appointment, street-turn, weather, congestion, or merchant-release records are absent. Transfer to another port or pool is limited unless definitions and event capture are comparable. The bounded conclusion is that utilization research is strongest when equipment identity, state transitions, and exposure time are joined before a benchmark is presented.