FIELD RECORD: The departure board said Wait. The loudspeaker said Go.
This is original systems fiction set in Clandestinia, not a report about a real transport service. The exercise uses invented journeys and is not an accessibility compliance certification.
The Call Above the Concourse
The morning after The Next Person at the Terminal, Mara returned to find the repaired machine asking each visitor to begin a new session. Across the concourse, a traveler named Iven was still watching the departure board.
His journey showed a blue rectangle and the word Wait. Above him, a loudspeaker called his boarding group to a different gate. Iven was deaf. The instruction never appeared on the board.
By the time he reached an attendant, his group had departed. The attendant opened an event log. Beside the instruction were two words: Announcement successful.
Iven placed his journey card on the counter. “Successful where?”
What the Log Could Prove
The log recorded that an audio file had played through the station’s announcement controller. It did not record a visual update, a personal notification or an acknowledgment from Iven. The attendant had mistaken evidence about a speaker for evidence about a traveler.
Mara asked the dispatcher to reconstruct the instruction. It contained three changes: boarding had started, the gate had changed, and the previous waiting instruction no longer applied. None had reached the display Iven was using.
The two channels were maintained separately. A dispatcher could interrupt the audio schedule, but the public board required another action. During the disruption, that action had been missed.
The station could explain how the mismatch happened. That explanation did not get Iven onto a train.
A Useful Alternative
The dispatcher arranged a new departure and gave Iven a written gate instruction before opening a repair ticket. The incident record described a failed notification route, not a traveler who had ignored a call.
Someone proposed displaying every passenger’s name on the main board. Iven pointed to the strangers around him. He needed the instruction, not a public account of his journey.
For the trial, the station used a boarding-group label already printed on journey cards. The display named the group, the action and the gate. It used words as well as color, and kept a recent-changes list so a brief glance away did not erase the instruction. Individual exceptions remained available through a private, accessible contact route.
The team also removed the smartphone requirement from its proposed fallback. A phone alert could be useful, but it could not be the only alternative for someone without a charged, connected device.
Iven helped review the revised journey and was paid for his time. His feedback exposed problems the dispatcher’s simulation had missed, including a gate abbreviation that made sense only to staff. He was one reviewer, not a representative of every access need.
Field assessment: a message can leave a system successfully and still fail to give its intended recipient a usable next step.
Review Exercise: A Journey Without Sound
Use a paper model or an authorized staging interface. Create a fictional journey with a boarding group, an initial gate and a later gate change. Do not disable live announcements or use real passenger information.
Write the instruction first. Specify the action, location and relevant timing. Then compare what each channel communicates. A vibration that merely says something happened is not equivalent to a readable gate-change instruction.
Remove assumptions. Walk through the journey without sound, without a smartphone, and after missing the first notification. These checks reveal design dependencies; they do not simulate the full experience of disability or replace testing with people who have relevant access needs.
Check privacy and meaning together. Decide which information belongs on a public board and which needs a private route. Use understandable labels and text rather than color alone. Make sure the alternative communicates the same actionable change without exposing unnecessary personal details.
Make disagreement visible. Give the audio channel a new gate while leaving the display on the old one. Record how staff detect the conflict, which instruction is authoritative, and how affected travelers receive help. Keep transmission, availability and acknowledgment as separate observations.
Record the unresolved cases and who owns the next test. Continue with the Digital Systems Accountability guide to connect a failed instruction to a correction route rather than another reassuring status label.