When a Town's Traffic Light Timing Becomes a Physics Problem You Can Solve

You've been sitting at the same red light for what feels like forever. The cross street got green, then yellow, then red, and you're still waiting. It's easy to blame the city or some bureaucrat behind a desk. But the truth is, traffic light timing is a physics problem—and you can solve it with a stopwatch and a few equations.

I'm not talking about hacking the controller. I mean auditing whether the yellow interval gives drivers enough time to stop. That's a measurable thing. And if it's off, the result is either rear-end crashes or frustrated drivers running reds. Either way, physics is your friend here.

Who Gets Stuck and Why the Default Timing Fails

The driver who hits red after red

You know the feeling. You roll through an intersection at 35 mph, the light turns yellow, and you brake—hard. Then green. Then red again at the next block. The whole corridor punishes you. Default traffic light timings are set by municipal software that assumes every car arrives at the same speed, the same weight, the same acceleration curve. That assumption is a lie. I once watched a delivery van trigger three consecutive reds on a suburban arterial because the timing assumed a 0.3 g deceleration rate, but the van’s worn brakes and 2,000 pounds of cargo pushed that to 0.22 g. The controller never knew. It just saw a vehicle approaching too fast, slammed the red phase early, and trapped everyone behind it.

The catch is—cities don't adjust for real driver behavior. They plug in a speed limit number, not the actual speed people drive. Wrong order. A driver who coasts at 40 mph where the plan says 30 will hit a yellow that decays into red before they can commit. The physics of stopping distance—mass, friction, reaction time—is entirely absent from the default spreadsheet. You lose two minutes per intersection. Over a week, that adds up to a half-hour of idling. Not catastrophic. But it gnaws.

The pedestrian who barely crosses

That crosswalk button you mash? It does almost nothing for timing. Most pedestrian phases are calculated for a 4 ft/s walking speed—the federal standard from the 1970s. But try crossing a 60-foot intersection with a shopping cart, a toddler, or a cane. The walk interval expires before you reach the median. You jog or you wait. This isn't a complaint—it's a geometry flaw. The pedestrian's traversal time is a simple equation: distance ÷ speed. If the city never measures actual foot traffic speeds, the timing fails the old, the young, and anyone carrying groceries.

I once timed a downtown crosswalk where the walk phase gave 12 seconds for a 72-foot crossing. That demands 6 ft/s—a brisk jog for most adults. The city engineer later admitted they used the default value from a 1974 manual.

'We never changed it because nobody complained in writing.'

— Anonymous traffic engineer, municipal workshop, 2023

The fix isn't hard: measure the slowest regular user. But the default timing treats pedestrians as theoretical dots, not heavy bags and tired legs. That hurts.

The city engineer buried in complaints

Here's the trade-off most people miss: engineers have to balance vehicle throughput against pedestrian safety against transit priority against emergency preemption. One variable shifts and everything cascades. A longer green for cars means a shorter crosswalk phase. A protected left turn steals seconds from the through lane. The default timing acts like these are equal weights—but a semi-truck needs 40% more time to clear the box than a sedan. Physics doesn't negotiate.

What usually breaks first is the coordination pattern. A corridor timed for 35 mph will fail at 45 mph or 25 mph. And that is why the engineer's phone rings every Tuesday at 5:15 PM. Citizens describe the same pain: "I hit every red." The engineer checks the plan—looks fine on paper. The paper doesn't include rain, potholes, or the dump truck that downshifts early. So complaints pile up, the engineer tweaks the green split by 2 seconds, and the problem moves one block west. The underlying physics never gets fixed. That's where you come in—because you can measure what the software guesswork misses.

Honestly — most physics posts skip this.

What You Need to Know Before You Start Timing

Reaction Time vs. Braking Time — They Are Not the Same Thing

The yellow light isn't a timer for when you should stop. It's a window for when you can. That distinction kills most attempts to fix intersection timing. People sit in their car and guess: "Yellow lasted about four seconds." What they actually measured was their own panic plus the car's actual deceleration plus the guy behind them honking. Useless data. You need to split the yellow interval into two physics pieces: the time it takes you to notice the light and move your foot (reaction time), and the time it takes the car to actually shed speed (braking time). Reaction time for a sober, alert driver sits around 1.0 to 1.5 seconds — that's not a guess, it's the standard from traffic engineering manuals. Add a phone glance or rain fatigue? You're pushing 2.0 seconds easily. That sound you hear is the entire safety margin evaporating before your foot even touches the pedal.

Stopping Distance: The Equation Without the Scary Symbols

You don't need calculus to figure out if a yellow light cheats you. You need three numbers: your speed in feet per second (multiply mph by 1.47), your reaction distance (speed × reaction time), and your braking distance. Braking distance is where most people guess wrong. A car with good tires on dry asphalt decelerates at about 10 to 15 feet per second squared. That's a 0.7 g stop — firm but not wreck-your-face-on-the-steering-wheel firm. Trucks and loaded vans? They're closer to 6 to 8 feet per second squared. The catch is that yellow lights in most towns are timed for the 85th-percentile speed — the speed 85% of traffic drives, not the posted limit. If that speed is 40 mph and a driver reacts in 1.5 seconds, their total stopping distance is roughly 135 feet. If your measured yellow duration only covers 120 feet of travel at that speed, you've found the flaw. Not a theory. A number your town council can't ignore.

“The yellow interval should let a driver either stop safely or clear the intersection. If it does neither, the timing is not a suggestion — it's a trap.”

— from a traffic engineer's field notes, shared during a city council workshop I attended last year

Typical Deceleration Rates — What Default Timings Assume

Most municipal timing formulas assume a deceleration rate of 10 feet per second squared. That's fine for a compact car with new tires and a driver who sees the light turn yellow from 300 feet away. What breaks first is the assumption that every vehicle at the intersection matches that profile. A Ford F-250 towing a trailer? That rig decelerates at maybe 7 ft/s² — its stopping distance is 40% longer. I have watched a loaded delivery truck blow through a yellow that was perfectly timed for sedans. The driver wasn't reckless. The physics simply didn't fit the number someone typed into a spreadsheet eight years earlier. That's the trade-off: you either time for the fastest, lightest vehicle and accept that trucks will run reds, or you time for the heavy vehicles and accept that cars will sit through a longer yellow cursing the city. The trick is finding out which scenario your intersection actually lives in. Go measure.

How to Measure Intersection Timing in the Field

Measuring green and yellow durations with a stopwatch

You pull up twenty feet from the stop bar, engine idling, phone timer ready. Start it the instant the light goes from red to green. Stop it when it flicks to yellow. Run this three times, not once—the first measured green was forty-seven seconds; the second read forty-nine. The third? Forty-eight. That split matters because a single stray data point hides the real story. Most drivers guess the cycle in their head and guess wrong. Stopwatch evidence doesn’t lie. Once you have the green length, reset and catch the yellow alone—it usually runs three to five seconds, but the variation between intersections is brutal. One suburban light near me had a yellow that lasted exactly 3.1 seconds, measured across five cycles. That's an eternity for a sedan and a death sentence for a loaded box truck. The trick: stand clear of the crosswalk so your phone stays steady and your thumb hits pause clean. Jab it early and the yellow appears a half-second shorter than it's.

Estimating approach speed with a GPS app or radar gun

You need the speed cars actually carry, not the posted limit. A GPS app like GPS Status or even a basic cycling computer gives you real-time velocity in tenths of a mile per hour. Drive the approach yourself at the traffic flow—match the pack, don’t lead it. Record your peak speed as you cross the detection zone, about fifty to seventy feet before the stop bar. Worth flagging—radar guns are overkill. They require practice, calibration, and a partner who won’t look like a cop. GPS is good enough. The catch: cheap phone apps lag by one to two seconds, so your reading under-reports speed if you accelerate hard. Smooth out the throttle and hold steady. I have seen a driver reading 32 mph on a straight stretch where the limit was 35, but the group behind him was flowing at 39. That six-mph gap shifts the stopping distance by nearly a full car length.

Repeat for ten vehicles in light traffic. Take the 85th percentile speed—that’s the number engineers use, not the average. If nine cars do 38 mph and one does 45, the 85th percentile sits around 42 mph. That's your working number. Ignore outliers: the guy doing 55 in a 35 is a cop magnet, not a physics reference.

“The yellow time equation is dead simple. Physics doesn’t care about your opinion of the city planner.”

— overheard from a traffic engineer during a town hall meeting, after a citizen insisted the light ‘felt too short.’

Calculating required yellow time using distance and deceleration

Here is where the stopwatch meets the calculator. Required yellow time = (approach speed in feet per second / deceleration rate) + reaction time. Deceleration rate for a dry road is about 10–12 ft/s² for passenger cars. That sounds fine until you realize a wet road cuts that to 7–8 ft/s². Plug 40 mph (58.7 ft/s) into the formula: 58.7 ÷ 10 = 5.87 seconds. Add a one-second reaction lag. You need 6.87 seconds of yellow. Most municipalities run yellows at 4.0 to 5.5 seconds. That gap—nearly two seconds—is the space where red-light-running happens. What usually breaks first is the assumption that everyone reacts in one second. Older drivers, distracted drivers, or drivers towing a trailer often need 1.5 seconds. That ripples the calculation up to 7.37 seconds. Not yet. You also have to account for the dilemma zone—the stretch where a driver can neither stop safely nor clear the intersection before the light turns red. This is where your field data matters. If your measured yellow time falls short of what the formula demands, you have the numbers to prove the timing is a physics problem, not a subjective gripe. Take your spreadsheet to the public works department. Show them the stopwatch logs, the GPS speeds, the deceleration math. Then ask them to recalculate.

Tools You Can Actually Use (No Lab Coat Required)

Stopwatch or phone timer

Your phone already has a stopwatch app. Use it. The built-in Clock app on iOS or Android samples time to 0.01 seconds—more precision than you need. The trap? You hit 'start' late, or 'stop' early. Reaction delay averages 0.2–0.3 seconds per tap. That adds up fast. I have seen people record a green-light window as 22 seconds when the real span was closer to 19. The fix is boring but critical: start the timer on the first visible movement of the opposing car, not when you notice it. Same for stopping—cut at the moment the last car clears the intersection, not when your brain registers 'clear'. Wrong order. One late tap and your entire data set drifts.

Odd bit about physics: the dull step fails first.

A cheaper alternative? Any digital wristwatch with a lap-split function. The Casio F-91W (ten bucks) does this perfectly. No battery anxiety—it runs for years. The catch is manual logging: you read the split, write it down, reset. That produces gaps. Phone timers let you screenshot the lap list, which saves time in the field. Pick the tool you will actually carry. Most people overthink this.

GPS speed app vs. laser rangefinder

You don't need a radar gun to measure vehicle speed through an intersection. A GPS speed app like SpeedView (Android) or GPS Speedometer (iOS) samples at 1 Hz or better—adequate for timing a 40-foot car crossing a 60-foot box. Accuracy hovers around ±1 mph at steady speed. That sounds fine until you hit stop-and-go creeping near the limit line; GPS lags by about a second, so instantaneous speed reads low. Worth flagging—if you're timing a truck accelerating from a stop, the app will show 15 mph when the vehicle is already doing 20. That skews your time-to-clear calculation by half a second.

A laser rangefinder (the kind golfers use, $150–$300) gives distance to ±1 yard. Combined with a stopwatch, you can compute speed manually: distance ÷ time = speed. The trade-off is setup time—you need a fixed point to bounce the laser off, like a reflector on a signpost. Not practical for every corner. I lean on the GPS app for quick scans and the laser only when I suspect the app is lying (wet roads, tree canopy, tunnel approaches). Most teams skip this: they trust the phone blindly. Check it once against a measured 100-foot mark. If the app reads 33 mph over a known 30-mph stretch, adjust or swap tools.

Notebook and spreadsheet for recording

A paper notebook survives dead batteries. Field notes are raw—time of day, weather, traffic direction, cycle phase lengths. Use a simple grid: column for start time, green duration, yellow duration, red duration, notes. That's three laps per cycle. After twenty cycles you have a pattern. The spreadsheet (Google Sheets or Excel) comes later: average the green times, spot outliers (one 45-second green in a row of 30-second greens—likely a pedestrian call or bus priority). That hurts—one stray data point can shift your average by two seconds if you don't flag it.

Rhetorical question: how many cycles do you need before the numbers stabilize? For a standard four-phase intersection, forty complete cycles (about two hours of observation) cuts the margin of error below ±1 second. Less if traffic is light. More if you're measuring during a holiday or construction. Log your start and end times religiously—I once spent an hour recording before I realized my phone had jumped time zones. The data was useless. Keep a pen that writes on wet paper; a Rite in the Rain notebook costs $12 and saves that headache.

'The difference between a hunch and a data point is a notebook you actually carried to the curb.'

— note taped to my dashboard after I forgot my recorder three days straight

What Changes When the Road Is Wet or the Truck Is Heavy

Reduced friction and longer braking distances

Dry asphalt gives you a coefficient of friction around 0.7 to 0.8. That’s the number that lets a typical sedan stop from 30 mph in roughly 45 feet. Now add rain—and that coefficient drops to 0.4 or worse. Suddenly your stopping distance stretches past 70 feet. The catch is that traffic-light timers don’t know it’s raining. They compute yellow intervals assuming dry pavement, which means the driver who started braking at the "safe" 3-second mark now skids through the stop bar. I have seen this exact failure at a four-way in Portland; the city had to reprogram the controller after three rear-end collisions in one wet week. The fix in your audit: measure the dry-braking distance, then multiply by 1.6 for wet conditions. If that extra length pushes the vehicle past the intersection before the light turns red, your town’s timing is wrong.

Vehicle weight and load effects

Heavier vehicles need more road to stop—but not for the reason most people guess. Friction scales with weight, yes, so a loaded delivery truck has more grip than an empty one. The problem is kinetic energy. That energy grows with the square of speed but linearly with mass. Double the mass, double the energy the brakes must shed. A dump truck hauling gravel can weigh 80,000 pounds; a passenger car weighs about 4,000. Same speed, twenty times the kinetic energy. Worth flagging—the brakes themselves overheat, too, which reduces stopping power after repeated use, a phenomenon called brake fade. So when you time an intersection, do it during a quiet period, then separately clock a heavy vehicle (bus, garbage truck, delivery van) through the same zone. The difference is rarely subtle.

Grade (uphill/downhill) impact on stopping

A 3% downhill grade adds roughly 15–20 feet to a 30-mph stop. That’s one full car length—enough to trigger the red-light camera on a short yellow phase. The physics is simple: gravity works against you on the downslope, adding a forward component to the vehicle’s inertia. Uphill does the opposite; you stop shorter, which means the intersection might actually have too much yellow time for that approach. Most teams skip this. They time the flattest lane and assume it represents all lanes. Wrong order. You need to measure each approach’s slope with a phone inclinometer (free, accurate to 0.1 degrees) and recalculate the stopping distance using the grade-adjusted friction formula: effective friction = coefficient × cos(angle) ± sin(angle). That sounds like math homework, I know—but one spreadsheet cell saves you from recommending a timing that fails on the hill.

Field note: physics plans crack at handoff.

‘We timed the flat road and called it done. First rain, first downhill truck, first red-light ticket that got thrown out in court.’

— Traffic engineer explaining why their audit had to be redone, as told by a city planner I work with

One more variable people forget

Tire condition. Not every car rolling through your intersection has tread depth above 2/32 of an inch—below that, hydroplaning starts at 35 mph on wet roads. Your audit can’t measure every tire. But you can build a safety margin: add 0.5 seconds to the yellow interval for any approach where you observe heavy truck traffic or a downhill grade. That half-second feels tiny until it prevents a 12-ton bus from blocking the crosswalk. The trade-off is that longer yellows make impatient drivers treat them as "more time to beat the red." That hurts compliance. So you balance: physics says one thing, human behavior says another. The best fix I have found is to adjust the all-red clearance interval instead—gives heavy vehicles time to clear without encouraging speed-ups. Your audit should recommend that, not just yellower yellows.

Why Your Numbers Might Be Wrong and What to Fix

Confusing cycle length with green time

The most common mistake I have seen in citizen audits is treating the full cycle as the green window. You stand at the curb, watch the light go red, then green, then red again, and you jot down “90 seconds” as your green time. Wrong order. That 90 seconds includes the red phase, the yellow transition, and the all-clear buffer. Your actual green time might be only 38 seconds—less than half. The catch is that traffic engineers design cycles around the *total* loop, not the green slice. If you present a cycle length as usable green, you will claim the intersection can move 40 cars when it can barely handle 18. I once watched a neighborhood group fight for a longer green based on a 120-second cycle. They were furious when the city showed them the green was actually 45 seconds. They had been standing through two reds without realizing it.

To fix this: use a stopwatch with a lap function. Start the watch when your side turns green. Stop it when it turns yellow. That's your green time. Let the watch keep running through the red—that gives you the full cycle. Two numbers, not one. Write them separately. If you mix them up, your entire audit is a mirage.

Using speedometer instead of GPS

Your car's speedometer lies to you. Not out of malice—manufacturers deliberately calibrate them to read 2-5 mph high so you never accidentally speed. That sounds fine until you use that number to calculate how long it takes to clear an intersection. A trucker who thinks he is doing 35 mph is actually crawling at 31. Over a 300-foot crossing, that error adds nearly two seconds to your timing estimate. Two seconds may not sound like much. It's the difference between a green light and a red-light ticket—or between a safe pass and a rear-end collision.

Most teams skip this: they check their phone GPS instead. Even then, average GPS accuracy on consumer phones is ±16 feet under trees or near buildings. City centers kill your signal. I have watched someone stand under a steel bus shelter, watching their app flicker between 28 and 44 mph while the bus sat still. That hurts. The fix is simple: use a dedicated GPS logger or at least cross-check with two apps on separate phones. If they disagree by more than 2 mph, re-run the measurement. Your numbers are only as good as your reference frame.

Forgetting the driver's reaction distance

“The light turns green. Your foot lifts. The car behind you doesn't move for a full second. That lost time compounds across every cycle.”

— field note from a traffic audit in a rain-soaked suburb, where reaction lag added 12 seconds per hour of delay

The human factor is the one variable you can't calibrate with a stopwatch. Reaction distance—the gap between “go” and “actually going”—varies wildly. A tired commuter staring at their phone might take 1.8 seconds to move. A seasoned driver watching the crosswalk might react in 0.6 seconds. That 1.2-second difference per car means a queue of 15 cars loses 18 seconds before the last car even enters the intersection. The bolt you tighten is not the light timing—it's the human bottleneck.

To double-check: during your next field session, note whether the third car in line moves within one second of the first. If not, your calculated throughput is optimistic. Adjust your green-time estimate by adding 0.8 seconds per vehicle for standard conditions, 1.5 seconds for wet pavement. Present those adjusted numbers to the city, not the raw stopwatch data. They will take you more seriously when your model matches what actually happens at 5:15 PM in a drizzle.