Critical-incident interview
Reconstruct a recent time-critical journey and the information that shaped each decision.
UX RESEARCH · PRODUCT DESIGN · 2026
WaySure
Clearer travel decisions, even when live data is not.
A decision-support concept for university students using Edinburgh buses and trams under time pressure.
ARRIVE BY 09:45 · ILLUSTRATIVE
Three good options
Compare prediction ranges and evidence.
414 min
09:35–09:40Likely on time · 5–10 min buffer624 min
09:33–09:48More uncertain · one transfer244 min
09:38–09:44Likely on time · less walkingRoleProduct designer & researcher
ScopeHCI coursework · Mobile experience
DeliverableResearch plan + high-fidelity prototype
ToolsFigma · Figma Make
01 / CONTEXT
When a precise time creates false confidence.
Real-time public transport information cannot always be completely accurate. Yet a student still has to decide: leave now, keep waiting, walk to another stop—or risk being late.
WaySure reframes the interface from a timetable into decision support. Instead of hiding uncertainty behind one exact number, it shows a range, the evidence behind it and the consequences of each next action.
02 / RESEARCH PLAN
Say · Do · Make
One 50–60 minute session combines three complementary methods. This is the planned research design; participant sessions and findings are not yet claimed as complete.
Think-aloud task
Plan an arrival, then respond when a selected stop closes but the app has not updated.
Journey Confidence Map
Map confidence, risk and desired support from planning through arrival.
03 / DESIGN HYPOTHESES
From more information to better decisions.
Make uncertainty visible
Use arrival ranges and confidence language rather than a single definitive minute.
Show the evidence
Expose freshness, vehicle tracking and stop status so trust is inspectable.
Compare consequences
Compare waiting, walking and rerouting by arrival range, effort and reliability.
A journey designed around decision moments.
01Set arrival goal02Compare routes03Inspect evidence04Follow journey05Detect conflict06Compare next steps07Recover08Arrive
Three moments where the interface earns trust.
The visual system combines a warm editorial surface, electric blue actions, coral disruption states and restrained hand-drawn annotations.
01
Compare realistic arrival ranges
Routes are ranked by expected arrival, but each option communicates its range, walking effort, transfers and likely buffer—not only speed.
ARRIVE BY 09:45 · ILLUSTRATIVE
Three good options
Compare prediction ranges and evidence.
414 min
09:35–09:40Likely on time · 5–10 min buffer624 min
09:33–09:48More uncertain · one transfer244 min
09:38–09:44Likely on time · less walkingRECOMMENDED · ARRIVE BY 09:45
09:35–09:40
● Likely on time · 5–10 min bufferLive vehicle tracked · updated 32 seconds ago
- Walk to South Bridge
- Wait at South Bridge
- Take the 41 to Old College
- Final walk to destination
Why this estimate?See the signals behind this range →
02
Explain a conflict before asking for action
When physical stop information and live predictions disagree, the interface names the conflict, downgrades confidence and shows exactly why the original estimate is unreliable.
! Information conflictingApp data and stop status disagree
ARRIVE BY 09:45
South Bridge stop may be temporarily closed.
The arrival prediction is not confirmed.
CURRENT ARRIVAL RANGE09:36–09:41
CONFLICTING EVIDENCE
CONFLICTING EVIDENCE
If this stop is closed, this range is unreliable.
Why trust is limited• Stop status unconfirmed• Live vehicle data has changed
03
Recommend a recovery, not just an alert
Wait, walk and reroute are presented in the same decision frame. The recommendation is tied to reliability and the chance of arriving on time.
ARRIVE BY 09:45 · COMPARE CONSEQUENCES
Compare wait, walk and reroute
Choose using reliability, effort and arrival range.
HIGH UNCERTAINTYWait here09:38–09:55Stop unconfirmed
MOST RELIABLE · RECOMMENDEDWalk 6 min to Chambers Street09:39–09:44Likely on time
MEDIUM UNCERTAINTYReroute09:48–09:53Two stops away
06 / REFLECTION
What this project demonstrates—and what comes next.
WaySure demonstrates how uncertainty can become a usable interface material rather than a hidden technical limitation.
The current result is a research plan and a high-fidelity design hypothesis. The next step is to run the planned sessions with 3–5 university students, analyse decision episodes across Say, Do and Make evidence, and revise the prototype based on observed behaviour.
- Test whether arrival ranges are understood quickly under time pressure.
- Check whether evidence improves trust without increasing cognitive load.
- Evaluate whether recovery comparisons lead to confident next actions.