Turning Movement Counts: Method and How to Read the Results

A classified volume count tells you how many vehicles of each type passed a point. A turning movement count (TMC) tells you where they went. At a junction, that difference decides signal timings, lane allocation and whether a proposed layout will work.

What a TMC measures

At a four-arm junction, every vehicle arriving on an approach does one of three things: turns left, goes straight, or turns right. Each approach therefore produces three movements, and a four-arm junction produces twelve — plus U-turns where they are permitted, and pedestrian crossings where they are counted.

A TMC records each of those movements separately, usually broken down by vehicle class and by time interval, most often fifteen minutes.

How the data is collected

Manually, with enumerators at each approach using counters or tally sheets. One person can reliably track a limited number of movements, so a busy junction needs several, and fatigue sets in over a long count.

From video, with cameras covering each approach. The footage is then processed — by people or by software — to produce the same movement-by-movement table. The advantage is that the record is reviewable: a disputed figure can be checked against the footage rather than argued about.

Either way, camera or enumerator placement determines data quality. An approach viewed at too shallow an angle, or a lane hidden behind a bus, produces gaps no amount of processing will recover.

How the results are reported

A completed TMC usually contains:

• A movement matrix — every approach against every exit, for the count period.
• Fifteen-minute interval data, which is what peak-hour analysis needs.
• Class breakdown per movement, because a junction carrying many heavy vehicles behaves differently from one carrying mostly two-wheelers.
• Peak-hour volume — the highest consecutive four fifteen-minute intervals.
• Peak-hour factor — the peak-hour volume divided by four times the busiest fifteen-minute interval. It describes how evenly traffic arrives within the peak: a value near 1.0 means a steady flow, while a lower value means sharp surges that the junction must absorb.

What planners do with it

• Signal design. Green time is allocated per phase from the movements that actually use it. A right-turn movement that is heavy for ten minutes and empty for fifty needs different treatment from one that is steady.
• Capacity assessment. Approach volumes against capacity show which arm fails first.
• Layout decisions. Whether a dedicated turning lane is justified, and how long it needs to be.
• Scheme appraisal. Before-and-after counts show whether a change achieved what it was meant to.

Common pitfalls

• Counting on an unrepresentative day. A market day, a festival, an exam day or rain produces a pattern the junction does not see the rest of the year. Record the conditions with the data.
• Too short a count. A two-hour count catches one peak. Understanding a junction usually needs both peaks and an off-peak period, and for some schemes a full day.
• Ignoring the mix. In mixed Indian traffic, converting everything to passenger car units too early hides what is actually happening at the stop line.
• No record to go back to. When a figure is challenged months later during scheme approval, a video-backed count can be reviewed. A tally sheet cannot.

How SurveySense handles TMCs

SurveySense processes junction footage into movement-by-movement counts with class breakdown and fifteen-minute intervals, so the survey output and the source footage stay linked. Where a count is later questioned, the frames behind the number are still there to check.