Converting a one-day vehicle count into a signal warrant study

A planner calls asking whether an intersection "needs a light." You pull up the link in your GIS and there's no count on file, no loop detector, nothing from the last ARAN run that covers that approach. You have one day of volume data, maybe from a quick tube count or a one-off turning movement count. The question is whether that's enough to run a warrant analysis, and which warrants it can carry.
What the MUTCD asks for
Most of the nine signal warrants in the MUTCD (2009 edition, Chapter 4C, and carried through into state supplements) are built around sustained volume over a representative day. Warrant 1 (Eight-Hour Vehicular Volume) wants eight hours of major- and minor-street volume that clear the 80 or 100 vph thresholds for the applicable condition. Warrant 2 (Four-Hour Vehicular Volume) wants four hours plotted against the curve in Figure 4C-1. Those two assume you have a full day of counts, typically 24 hours split into 15-minute bins, so you can pick the four or eight hours that qualify.
Warrant 3, the Peak Hour warrant, works differently. It's built around a single hour: either the adjacent land use's peak demand window (a school, a plant shift change) or the day's single highest-volume period. If your one-day count captures that window cleanly, with 15-minute resolution so you can compute a real peak hour factor, you can run Warrant 3 off that single day. This is usually the fastest path for a planner who needs a preliminary answer before committing to a full warrant package for a capital project.
Warrants 4 through 9 (pedestrian volume, school crossing, coordinated signal system, crash experience, roadway network, and the two combination warrants) each have their own data requirements, and a few of them, like crash experience, don't depend on a vehicle count at all.
Using one day to screen, not to close the file
A single day tells you what happened on that day, nothing more. Traffic volume on a given corridor moves with day of week, with season, with whatever's happening at the adjacent land use that week. A Tuesday count in October doesn't automatically represent a Monday in June, and MUTCD guidance (and most state DOT manuals) expects you to account for that before a warrant analysis goes into a staff report as final.
In practice, this works two ways. If the one-day volumes clearly fail every applicable warrant by a wide margin, that's usually enough to close the question. Document it and move on. If the volumes clearly pass by a wide margin, same thing in reverse: you've got your answer. The gray zone is where a count sits close to a threshold. There, apply a seasonal adjustment factor from a nearby permanent count station, or at minimum flag the count as a screening-level result and note that a second day or a different season should confirm it before a signal gets designed off it.
Either way, what you need from the count itself doesn't change: volumes by movement and direction, in enough resolution to find both the warranted hour(s) and the peak hour factor, for both the major and minor street.
Getting the count without a crew
Getting a crew out to the link to run the count is usually what eats a week of schedule. The warrant math itself takes an afternoon. Tube counts need lane closures or at least a safe shoulder to lay the hoses, and a lot of the links that come up in these requests are exactly the ones nobody's instrumented and nobody wants to send a crew to for a one-off study.
Vehicle Counting turns a scheduled satellite or aerial pass over the corridor into counts by link and direction, delivered as a CSV plus the marked-up frame, no tube count to lay and no lane closure to arrange. For a Warrant 3 screening or an initial read on whether a corridor is anywhere near warrant territory, that's often the fastest way to get a defensible number on the table.
If a link on your list has never had a count run on it, that's the gap worth closing first.