Reference
Notes on how road measurement is actually specified
Standards, class schedules and method notes for Indian survey practice. Written for engineers looking something up rather than for anyone being sold to — the figures are the same ones our own software applies, so you can check our output against them.
Classes
19
Measured against
16,560
hand-counted vehicles
The notes
Four references, no gate
Vehicle classes and PCU factors
All 19 canonical classes with the passenger car unit factor each carries, and a worked conversion from a published count.
The classified volume count, and what a DPR needs
What the deliverable has to contain, which parts recorded video can produce, and the three surveys it cannot.
Automated counting against manual counting
Manual counting is the reference we are measured against. Four cases where you should not buy this, and the variables that decide the cost.
Counting traffic from footage you already have
What makes an existing camera position usable, what pixel height does to accuracy, and why most corridor CCTV was not mounted for counting.
Why these exist
A number you cannot check is not a measurement
Every figure in these notes is the figure our software applies, not a rounded version of it. If a PCU total in one of our outputs does not match the schedule published here, one of the two is wrong and we would want to know which.
The same principle runs through how we measure our own error: both accuracy measures published, the stricter one headlined, and the conditions under which ours degrades stated rather than omitted.
Check it against your road
Send fifteen minutes of footage
We will return the classified count, the PCU summary and the annotated video, so every figure in these notes can be checked against the frame it came from.
Send us footage