What is a Turning Movement Count (TMC) and Why Does It Matter?

A Turning Movement Count (TMC) is a traffic survey that records the number of vehicles making each possible turning movement at an intersection during a defined time period. At a standard four-way intersection, this means counting vehicles going straight, turning left, turning right, and making U-turns for each of the four approaches.

Traffic engineers use TMC data to understand how traffic moves through intersections. This data directly informs intersection design decisions, traffic signal timing, road safety audits, and transport planning models. Without accurate TMC data, engineers are essentially guessing about how traffic behaves at a specific location.

Why TMC data matters

Signal timing is one of the most direct applications. Traffic signals are typically timed based on volume ratios between competing movements. If the north–south straight movement carries three times more vehicles than the east–west movement during the AM peak, the signal phases should reflect that. TMC data provides the volume ratios needed to optimise signal timing and reduce unnecessary delay.

Intersection geometry decisions — whether to add a dedicated left-turn lane, widen an approach, or convert a signalised intersection to a roundabout — all depend on TMC data. The Peak Hour Factor (PHF) is particularly important here: a PHF close to 1.0 means traffic is evenly distributed through the peak hour, while a low PHF indicates short surges that require the intersection to accommodate much higher peak volumes.

In South Africa, SANRAL and the South African Road Traffic Signs Manual (SARTSM) provide guidelines on TMC collection methodologies and reporting formats. Most transport planning models, including EMME and Cube, accept TMC data directly.

Traditional TMC collection methods

Historically, TMC surveys required field observers — one or more people standing at an intersection with mechanical tally counters, recording every vehicle movement by hand. This is time-consuming, expensive, prone to human error, and creates a safety risk for the observers. A 12-hour survey with four approaches typically requires at least two observers working in shifts.

Video-based counting was the next evolution: a camera is set up at the intersection, footage is recorded, and a team reviews the video manually in the office. This is safer and more accurate than field counting, but still labour-intensive.

AI-powered video analysis, like that used in TallyRoad, automates the review process entirely. The same video is processed by machine learning models that detect and track vehicles, then classify their movements based on which approach lines they cross.

The format of a TMC report

A standard TMC report includes an interval breakdown (typically 15 minutes), a TMC matrix showing movement volumes by approach, peak hour analysis including PHF, and a classification breakdown by vehicle type. The standard output format in South Africa is Excel, which feeds directly into transport planning workflows.