Collecting Public Input for a Corridor Study
Corridor studies turn on where people want to go, and travel-demand models are structurally blind to trips the current network makes impossible. Asking residents to draw those trips is how latent demand gets into the record.
What the model can't tell you
A travel-demand model is calibrated on trips people currently make. That makes it excellent at forecasting the effect of a change on existing behavior and fundamentally unable to see a trip that nobody makes today because the network makes it a ninety-minute ordeal with two transfers. Those suppressed trips are often the entire case for a corridor investment.
The only reliable way to find them is to ask, spatially and without leading. An open Route Creator question — "draw a trip you'd like to make but currently don't" — fielded before any alternatives are on the table produces a density pattern of desire lines, and that pattern regularly puts an alignment on the table that the technical process would not have generated on its own.
Ask open first, then ask about alternatives.
The order matters more than most study designs assume. Once you show people three alignments, you can no longer learn what they would have drawn — the alternatives anchor the answer.
Round one — before alternatives
A Route Creator question with no alignments shown: draw the trip you take today, and draw the trip you'd take if you could. Constrain the map to the study area with a Map Boundary. Add trip purpose and time of day. That's the whole instrument — four questions, ninety seconds.
Output: a desire-line density map that goes into alternatives generation.
Round two — at screening
One Route Display question per alternative, each with its own map and its own reaction question, so nobody is asked to parse three alignments on one map. Then a single Route Ranker across all of them, which forces the same trade-off the study has to make.
Output: a ranked preference with demographic and geographic cross-tabs.
Regional distribution gets you regional opinion.
A corridor study fielded through the agency's general channels produces responses from across the service area, most of them from people with no stake in this corridor. Distribution has to be geographically targeted or the result is unusable.
- Stops and vehicles along the corridor — posted QR codes, the highest-relevance channel available
- Businesses and institutions fronting it — flyers and window cards at the trip generators
- Adjacent-zip email — segment existing lists geographically before sending
- Kiosk at corridor open houses — walk-ins into the same dataset
- Map Boundary on the survey — drawn trips can't land outside the study area
Check the density map at the midpoint. If respondents aren't corridor-adjacent, the distribution needs fixing while there's still time.
Straight into the GIS
Every drawn and selected route exports as GeoJSON with full coordinates, paired to a CSV of the tabular answers. Overlay drawn desire lines against your alternatives in QGIS or ArcGIS Pro and the gap between what was proposed and what people wanted becomes a picture rather than an argument.
Common questions
- What does a corridor study get from a survey that it cannot get from modeling?
- The trips the model does not know about. A travel-demand model is calibrated on observed behavior, which means it is very good at telling you about trips people currently make and structurally blind to trips people would make if the network allowed it. Open route drawing surfaces exactly that latent demand — and it is frequently the argument that distinguishes between two alternatives the model scores as equivalent.
- How do we present multiple alternatives without confusing people?
- One Route Display question per alternative, each with its own map and its own reaction question, followed by a single Route Ranker across all of them. Showing three alignments on one map at once reliably produces confusion; showing them one at a time and then asking for a ranking does not. The ranking is also the output that maps most directly onto how the study actually has to decide.
- How do we reach people along the corridor rather than the whole region?
- Distribution does most of the work: QR codes at stops and on vehicles along the corridor, flyers at businesses and institutions fronting it, and email to any list you have with addresses in the adjacent zips. Constrain the map with a Map Boundary drawn around the study area so drawn trips stay inside it, and use the density map to check mid-study whether your respondents are actually corridor-adjacent.
- What format does the geometry come out in?
- GeoJSON, with full coordinates for every drawn or selected route, paired to a CSV of the tabular answers. It opens directly in QGIS, ArcGIS Pro, geopandas, and R's sf package. The GIS desk gets real geometry to overlay against the alternatives rather than a text column to parse.
- Does this work for rail and BRT studies as well as bus?
- Yes — the map question types are mode-agnostic. What changes with a fixed-guideway study is the emphasis: station and stop location questions matter far more than routing questions, because the alignment is usually constrained by right-of-way while station placement is genuinely open. Route Display for the alignment plus a selection question for candidate station areas covers it.
- When in the study should we field it?
- Twice, ideally. Once early with open route drawing and no alternatives shown, to find out where people actually want to go before the alternatives constrain the conversation; and once at alternatives screening with the Route Display and Ranker questions. The first round is what stops you from screening out an alignment that would have served a real market nobody at the table knew about.
Field round one before the alternatives get drawn
The open-drawing round is only useful before alternatives are public, and it's a four-question survey. Build it during a free trial and get it posted.
Start free 14-day trial Origin-destination software