A green wave depends on whose journey counts
Adaptive signals allocate shared time. Define the journeys an improvement measure covers, then check what happened to the other movements at the intersection.
A driver can reach a series of green lights while someone crossing the road waits longer. Both experiences belong in the evaluation of the signal timing.
A signal allocates time between movements that can conflict. Buses share the road with cars, while bicycles may need their own movement. People crossing on foot also need time within the applicable safety requirements.
Sensors can report changing demand for software to use. The city still needs to define which journeys it intends to improve. I'd want the resulting measure beside an account of what happened to other road users.
Choose an objective before measuring a saving
FHWA's adaptive-signal guidance links evaluation measures to operational objectives. Smooth flow needs an appropriate measure, while access equity asks another question. Queue management also differs from throughput. Demonstrating one objective leaves comparison with alternative systems to further evaluation. FHWA's evaluation guidance
A controller can improve its target measure while another movement receives the same service or experiences a longer wait. The assessment needs to report which effect occurred.
Imagine a corridor where a longer main-road green reduces its queue. Vehicles on the side street then wait longer. The timing calculation can be correct while the policy choice still depends on whose delay the city intended to reduce.
Report that trade-off beside the average. The selected measure should make clear which journeys enter the calculation.
Check what the detectors observe
A fixed plan assigns time by schedule. An actuated controller responds to detected demand. Adaptive control uses changing-condition data to adjust timing over a corridor or network.
FHWA's timing manual explains the dependence on reliable detection with ongoing maintenance. The control decision needs an observation system that represents the traffic the signal is intended to serve. The signal-timing explanation
A failed detector can remove a movement from the demand estimate. The controller may then allocate time using an inaccurate account of who is waiting. The resulting delay can persist despite the sophistication of the calculation.
I'd inspect how the installed equipment detects pedestrians. Bicycle detection needs examination too, alongside motor vehicles. The operating rules should state what happens when detection is unavailable.
Keep required intervals inside the design limits
My earlier note gave simple yellow durations with all-red durations as a general recipe. Those values require site-specific engineering.
The U.S. MUTCD's eleventh edition, Revision 1, is dated December 2025. It requires yellow duration to follow engineering practices, with an engineering basis for red clearance where used. Its pedestrian provisions distinguish the invitation to begin crossing from the interval provided to finish. The current MUTCD
Those provisions apply within their U.S. regulatory context. The broader design question is which safety requirements constrain the timing objective at the site being evaluated.
A travel-time target needs to be evaluated within those requirements. The operating design should specify which parts of the sequence can change, leaving required intervals protected by their own rules.
Adaptive control describes adjustment within that design. Its documentation needs to identify the scope of the adjustment so a reviewer can understand what the controller is permitted to change.
Include speed management in the stated purpose
Portland describes "rest in red" as a speed-management operation responsive to people walking. It also responds to people cycling or rolling. Some signals rest at red in light traffic. The city states that interrupting dangerous high-speed travel is part of the purpose. Portland's explanation
That objective can deliberately add a stop. The agency has chosen a different service goal for those conditions, which the evaluation needs to test on its own terms.
The city's explanation describes design intent. Estimating crashes prevented would require an outcome evaluation capable of supporting that estimate.
An explicit purpose lets the city explain the additional stop. Measured results can then show whether the operation achieved the intended effect.
Evaluate the journeys affected
State the location before defining the observation period. Describe demand with the comparison plan used as a baseline. An obsolete baseline may yield a different claimed improvement from a maintained conventional plan.
Choose measures that reflect the objective. Main-road travel time can miss side-street delay. Bus performance may change independently, while crossing experience needs its own assessment. Reliability also needs a definition tied to the service being evaluated.
The original percentage-saving claim lacked that baseline with its context. I removed it because a general figure could not explain what improvement a particular site should expect.
The signal system's objectives need to be explainable to street users. Detection supplies information for allocating time. An evaluation then needs to show what happened to the journeys the city intended to serve.