How City Traffic Lights Are Actually Timed: A Look Inside Signal Control

Stand at any major intersection long enough and you will notice a rhythm: a green wave that carries a platoon of cars through three lights in a row, or a left-turn arrow that only appears at certain hours. None of that is accidental. Behind every traffic signal is a carefully engineered timing plan — and, increasingly, several of them.

Traffic light timing is the quiet backbone of urban mobility. It decides who waits, who moves, and for how long. Understanding how cities set their signals explains a lot about the congestion we experience every day — and where the biggest opportunities for improvement lie.

The core unit: the cycle

A traffic signal operates in repeating cycles, each divided into phases — discrete periods during which a specific set of movements (through traffic, left turns, pedestrians) gets the right-of-way.

The three numbers that matter most:

• Cycle length — how long it takes the signal to run through all phases once. Typical urban cycles run 60 to 120 seconds.
• Green split — how much of the cycle each approach or movement receives.
• Offset — the time relationship between adjacent signals, which determines whether a driver hits greens or reds down a corridor.

Get these three right and traffic flows smoothly. Get them wrong and you get stop-and-go waves, idling, and frustrated drivers.

How the timing is decided

Cities generally use a few well-established approaches, often in combination:

1. Fixed-time (pre-timed) control. The signal runs a fixed cycle and split regardless of what's happening on the road. These plans are calculated from historical traffic counts and are simple, cheap, and predictable — but they can't react to live conditions.

2. Actuated control. Detectors (induction loops, radar, or cameras) sense vehicles waiting at an approach and extend or shorten the green accordingly. Actuated signals skip phases when nobody is waiting and hold the green for a busy movement — far more responsive than fixed time.

3. Time-of-day (TOD) plans. Instead of one plan all day, the signal switches between several pre-tuned plans across the day. A typical city intersection might run five or more plans: an AM peak, a midday, a PM peak, an evening, and an overnight "free" plan with very short cycles.

Time-of-day control: the workhorse of modern cities

Most urban signals today are time-of-day actuated: the controller holds a library of timing plans, each tuned to a distinct demand pattern, and a clock switches between them automatically.

• AM peak — green biased toward inbound traffic and school/work corridors.
• Midday — balanced splits, moderate cycle lengths.
• PM peak — green biased toward outbound and commercial corridors; often the longest cycle of the day.
• Night / overnight — short cycles, frequent green for the major road, sometimes flashing yellow on the minor road.
• Weekend and holiday plans — lighter, more balanced demand; many systems run a separate weekend schedule.

These schedules are typically engineered from manual traffic counts or sensor data collected over days or weeks, then refined through observation and citizen feedback. Once set, they may not be revisited for years.

The limits of the current approach

Time-of-day plans work well for predictable demand. But real traffic is messy:

• Incidents and events throw the schedule off — a crash or a stadium event can invalidate a plan within minutes.
• Plans go stale. As the city grows and travel patterns shift, a plan tuned three years ago may no longer match reality.
• Data collection is expensive. Manual counts are labor-intensive, and permanently installed detection is costly to build and maintain — so many signals run on outdated assumptions.

The result is the gap most of us feel: a green light with nobody in the lane, or a left-turn phase that backs up while through traffic idles.

Where this is heading

The next step is adaptive signal control — systems that continuously measure demand in real time and adjust cycle length, splits, and offsets on the fly, rather than switching between a handful of pre-set plans. Camera-based traffic intelligence is the key enabler here, feeding live vehicle counts and turning movements into the control system minute by minute.

For now, though, most of the world's intersections still run on the time-of-day playbook: a clock, a handful of carefully tuned plans, and the engineering judgment that went into them. It works — but the room to improve is enormous.