Every pavement design, every HPMS submission, and most freight and truck-route decisions start with the same question: how many of each kind of vehicle passed this point? The answer is written in a 13-bin code that every state DOT uses and almost nobody outside traffic engineering has heard of.
This guide explains the FHWA 13 vehicle classes, what each one looks like on the road, why agencies depend on the split, and what it takes to count it accurately.
The Federal Highway Administration's Traffic Monitoring Guide defines 13 vehicle classes, often called Scheme F. The scheme sorts vehicles by the number of axles, the number of units (a single frame versus a tractor and trailer), and tire configuration. It was designed so that a count taken in Oregon means the same thing as a count taken in Georgia, and so that the classes map to how much damage each vehicle does to pavement.
Classes 1 through 3 are light vehicles. Class 4 is buses. Classes 5 through 13 are the truck bins, and they are the ones that matter most for the decisions below.
Pavement design. A loaded Class 9 does thousands of times more damage to a pavement section than a passenger car. AASHTO's mechanistic-empirical design method takes the full truck class distribution as an input. Get the Class 9 and Class 5 counts wrong and the pavement is either over-built at taxpayer expense or under-built and failing early.
Federal reporting. The Highway Performance Monitoring System requires classification data on sample sections, and the Traffic Monitoring Guide sets the expectations for how it is collected. Those numbers flow into apportionment and performance measures.
Freight and truck routing. Which corridors carry doubles and triples, where the Class 9 volume is growing, and which local roads are seeing truck traffic they were never designed for. All of it starts with a classification count.
Safety and enforcement. Truck percentage feeds crash analysis, weigh station planning, and the case for truck parking along a corridor. Our post on traffic studies and what they are used for covers how classification fits into the broader study.
The scheme is built on axles. That is the root of most errors. A sensor that cannot see axles has to infer the class from length or silhouette, and the trucks that matter most, Classes 5 through 9, overlap heavily on length. A short box truck and a long pickup with a trailer can be the same length and belong in different bins.
The other failure modes are mundane. Pneumatic tubes get cut or kicked loose. Loops and piezos need a pavement cut and drift out of calibration. Camera-only systems lose axles in rain, glare, and at night. A 48-hour count that loses six hours to a weather gap is a count that gets redone. We go deeper on each technology in our comparison of classification sensors.
The current approach is to combine two kinds of sensing on one device. Radar measures the speed, range, and length of every vehicle in any weather, day or night. AI video sees the axles, the trailer gap, and the body type that length alone cannot resolve. Fusing the two on the device classifies each vehicle into its FHWA bin as it passes, without anything on the pavement.
That is how FusionSensor delivers 13-class counts. It fuses HD3D radar and HD video 20 times per second, classifies at 98 percent or better in field conditions, and streams the result in the formats agencies already use, including NTCIP and MS2, from a single 15-watt device. PennDOT added it to its qualified products list for exactly this kind of data collection.
The FHWA 13 classes are the common language of traffic data. Every downstream decision assumes the count behind them is right. If you are planning a classification study this season, start with our checklist for running one that holds up, or see how we approach traffic data collection without putting anything on the road.
Subscribe to The Edge Brief for monthly field notes, deployment stories, and product updates.
Subscribe →