Week 10 Field session: The post-mortem

Working through a published inquiry into a queue that failed badly, and finding the week we would have caught it.

12 May 2027

This session stays in the room, works on paper, and is a quiet one. You spend two hours with a document that somebody wrote after something went wrong.

What happens#

Each group takes one published report, which might be a crowd safety inquiry, a transport incident investigation, a health service capacity review or a post-incident report from a large system outage. All of them are public documents and all of them are about a queue.

Forty minutes, reading. Alone, not as a group. You read the findings section and the timeline.

Forty minutes, reconstructing. As a group, draw the queue. Servers, arrival rate, capacity, discipline, and the bound that was not there.

Thirty minutes. Two questions. Which single quantity, if somebody had been measuring it and looking at it, would have shown this coming. And which week of this course covers it.

Ten minutes. Each group answers the second question out loud. The answer is usually week 3 or week 10, and hearing twenty groups converge on that is the point of the exercise.

How we handle this#

Several of these documents concern deaths. We work from the engineering findings and the timeline, not from the accounts of the people involved, and we do not go looking for footage. If you would rather take a report without fatalities, two of the six are outages and nobody will ask you why.

Bring#

• Nothing. Reports are printed and provided.
• Read the week 10 lecture notes first if you were not in the room.

What leaves the room#

A reconstructed queue diagram for the incident, one named measurable quantity that nobody was watching, and the week of the course that covers it.