The point: There is a frustrating signal on Rt101 in Greer, SC that only gives green to the north or south approach one at a time. The 546 Junction Lab is a simulation of that intersection, so you can test whether a different timing plan would cut the waiting.
How to use the 546 Junction Lab Watch it run. Cars drive through the junction automatically. Use Pause/Play, the 1×/4×/12× speed buttons, and “Zoom to junction” to get a closer look. Cars shift from white to yellow to red the longer they’ve been delayed, and the readout under the picture shows average delay, cars stopped on each approach, and the current signal state.
Set the traffic (right panel). Pick Light, Typical, or Rush hour, or use the sliders for cars per hour from the north, south, and 546, plus how many turn. The starting numbers are placeholders, so replace them with counts you measure at the real light.
Set the signal plan. The default, “Today: one side at a time,” mimics the real light. Try “North + south together” or the preset chips for a longer main green, a south leading green, or sensor-actuated timing (which ends greens early when no cars are waiting). Sliders adjust green, yellow, and clearance times, and the timing diagram shows the cycle.
Test it fairly. Press “Test this plan.” It runs 15 minutes of traffic on three random patterns, the same for every plan, and adds a row to the comparison table with average delay per approach, longest queue, and cars served. Change something, test again, and compare. The lowest-delay plan gets a tag, and the × button removes a row.
Set the traffic (right panel). Pick Light, Typical, or Rush hour, or use the sliders for cars per hour from the north, south, and 546, plus how many turn. The starting numbers are placeholders, so replace them with counts you measure at the real light.
Set the signal plan. The default, “Today: one side at a time,” mimics the real light. Try “North + south together” or the preset chips for a longer main green, a south leading green, or sensor-actuated timing (which ends greens early when no cars are waiting). Sliders adjust green, yellow, and clearance times, and the timing diagram shows the cycle.
Test it fairly. Press “Test this plan.” It runs 15 minutes of traffic on three random patterns, the same for every plan, and adds a row to the comparison table with average delay per approach, longest queue, and cars served. Change something, test again, and compare. The lowest-delay plan gets a tag, and the × button removes a row.
Two limits to keep in mind: it’s one lane each way, and the gas station and diner driveways aren’t modeled, so treat results as a way to compare plans rather than exact predictions.
Full disclosure, I have no idea what I'm doing. I just threw this together with AI in 30 minutes. The intent is to bring awareness to the issue of how terrible this light is. I'm not sure of the accuracy of the timing or traffic flow.
Full disclosure, I have no idea what I'm doing. I just threw this together with AI in 30 minutes. The intent is to bring awareness to the issue of how terrible this light is. I'm not sure of the accuracy of the timing or traffic flow.
546 Junction Lab
Change the signal phases, timing and traffic, then test each plan against the same simulated cars.
Car colour by time lost0 s → 60 s+
- Avg delay, last 5 min
- –
- Cars stopped (N · S · 546)
- –
- Signal now
- –
Schematic of the intersection with one lane each way. Driveways at the gas station and diner are not modeled, and the starting numbers are estimates until real timing and traffic counts are entered.
Signal timing diagram
Compare plans
Each test runs 15 minutes of traffic after a 2-minute warm-up, on three different random traffic patterns, so every plan faces the same cars. Change a setting, then test again.
| Plan | Avg delay | North | South | 546 | Longest queue | Cars | Remove |
|---|