Walking a test route
You got a personal link like
…/field?t=yourname&day=2. It opens
today's assigned route pair on your phone. Compare the
three options and rate the choice, then walk your
assigned one, tap the numbered dots where something is
off, and rate the walk at the end. One walk per day,
8–20 minutes, inside your walk window. That's the
whole job.
Before you leave
- Open your personal link in your phone browser and allow location access when asked.
- The map shows your assigned route with a colored badge: fast, calm, or safe. That is the lens for today's walk — judge the route against that promise.
- Your assignment includes a walk window (e.g. "walk between 13:00–17:00"). The frozen route was computed for that time of day, so walk inside the window — otherwise what you see on the street won't match what the engine believed.
- No route or wrong name? Tell your field test manager — assignments are managed centrally.
Before you walk: compare the options
The app first shows all three options — Fast, Calm, and Safe — on the map, with time and distance chips. While they're side by side, rate choice value: were the options meaningfully different and worth choosing between (high / some / low)? Pick any issue chips that apply, add a note, save. Then the map switches to your assigned route only and the walk begins.
| Compare chip | Use when |
|---|---|
| Options too similar | The three lines are basically the same route — nothing to choose between. |
| Differences are the wrong kind | The options differ, but not in ways that match the Fast / Calm / Safe promises. |
| Benefit not worth the cost | The calmer or safer option exists, but the detour is too expensive for what it buys. |
| Better option missing | You know a route that should be one of the options and isn't. Dictate where it goes in the note — a reviewer will reconstruct it and add it to the bench. |
| Other | Anything else — explain in the note. |
During the walk
- The map keeps itself centered on you. Panning pauses that; the round button (bottom right) re-centers.
- The screen stays awake while the app is open. GPS stops when the phone is locked — that's fine, it re-acquires in a second or two when you pull the phone back out at the next decision point.
- GPS drifts near stations and tall buildings. Always tap the dot you mean — your blue position dot is a hint, not the anchor.
Annotating a node
Tap a numbered dot where something is wrong (or notably good — use the note). Pick the issues that apply, add a note, save. Fifteen seconds. The third column shows where each chip routes at review time — every chip terminates in exactly one fix surface.
| Issue chip | Use when | Where it goes |
|---|---|---|
| Couldn't walk it | Closed path, construction, private access, barrier, no sidewalk, impossible crossing. | Graph data fix |
| Path doesn't match reality | Route cuts through a building, uses a path that doesn't exist, misses one that does. | Graph data fix |
| Wrong entrance / access | Route starts, ends, or enters at the wrong place — wrong side of street, service entrance, wrong station exit. | Data or snapping — review disambiguates with street-level imagery |
| Confusing / unneeded instruction | The guidance here says something odd — a turn that isn't one, "slight left" on a straight path, too many instructions at once. | Guidance / instruction generation |
| My location looked wrong | Blue dot jumps to another street, lags, or shows you off-route while you're on it. | Frontend / GPS — excluded from routing triage |
| Loop / backtrack | Route loops, doubles back, or passes the destination and comes back. | Engine / graph |
| Time / distance look wrong | The ETA or distance is implausible for what you actually walked. | Speed model |
| Other | Anything else — explain in the note. | Sorted at review time |
On top of these, you get mode-specific chips — only the ones for the mode you're walking today:
| Mode | Issue chip | Use when | Where it goes |
|---|---|---|---|
| Calm | Traffic / noise / crowds here | This spot is loud, busy, or crowded — the opposite of the Calm promise. | Attribute gap — counted as a known gap and quantified, not as a profile bug |
| Calm | Quieter / greener street nearby missed | A calmer parallel street exists right there and the route ignored it. | Weights or data on existing dimensions |
| Safe | Feels dark / isolated / exposed here | This stretch feels unsafe — deserted, poorly lit, no one around. | Weights (lights / presence) or data |
| Safe | Stressful crossing / traffic decision | A crossing or junction here forces a stressful judgement call. | Attribute gap |
| Fast | Pointless detour here | The fast route wanders where a straight shot obviously exists. | Weights / topology |
- Notes: tap the text field and use the keyboard's mic key to dictate — faster than typing with cold hands.
- Photos: use your normal camera app (with location on), then tick "I took photos here". Photos are matched later by time and location — nothing to upload.
After the walk
Tap Finish walk and rate two rows — same scale the whole team uses. Choice value was already rated at the compare step, and mode alignment is rated only for the mode you actually walked.
| Rating | Question it answers |
|---|---|
| Route correctness | Does it work as a real pedestrian route — walkable, plausible, right start and end? |
| Mode alignment | Did the route deliver today's promise (Fast / Calm / Safe)? |
Ratings are health metrics, not verdicts — no single decision hangs on one. The chips and notes are what drives fixes. "Ratings tell us how bad it is; chips and joins tell us what to fix; the bench tells us whether we fixed it."
Good to know
- Dead spots are fine. Annotations queue on the phone ("N unsent" in the top bar) and send automatically when you're back online. Don't re-enter anything.
- What gets recorded: your issue picks, notes, ratings, and a single GPS snapshot per annotation as evidence. There is no continuous location tracking.
- Annotated dots turn green so you can see what you've covered.