Skip to main content
Marquee safety analysis specifies more checks than a crew can perform

Marquee safety analysis specifies more checks than a crew can perform

School of Continuous Improvement researchers enumerated 141 ways a folding event marquee can fail, ranked them on a pre-registered scale, and costed the schedule of documented controls the ranking generates at 214 checks per event.

A criticality analysis of 141 failure modes across 96 community events, and the control schedule the University has adopted from it

The School of Continuous Improvement has spent the year asking what its methods do when they are carried somewhere they were never designed for. A Criticality Analysis of the Folding Event Marquee, released this week, is the clearest answer it has produced.

The poster applies a failure mode, effects and criticality analysis — the apparatus written for ventilators and aircraft subsystems — end to end over the pop-up canopy that appears at fetes, sports days and market mornings. A six-member panel enumerated 141 failure modes across 47 marquees observed at 96 community events, scored each for severity, occurrence and detection, and ranked the result. Thirty-eight modes clear the action threshold. The highest-ranked of them is stowing the canopy damp, a fault nobody present at the time can see and which reaches a different crew a fortnight later.

The method does not know what it is looking at. Point it at a rocket motor or at a canopy on a school oval and it will give you the same kind of answer, at the same length.

— Dr Bram Ntuli, Lecturer and Convenor of Improvement Grand Rounds

The resulting schedule of documented controls runs to 214 checks per event, against a median volunteer base of nine, and has been taken up in full by the Indicator Commons. Dr Anneke Tolan, Senior Research Fellow in the School of Emergent Priorities, said the fieldwork “depended on crews who let us stand about with a clipboard while they were getting a stall up, and the School is grateful for their patience.”

The University regards the work as an early demonstration of what its assurance methods can be asked to carry. The full poster is available from the University’s research repository under an open licence, doi:10.5555/slop.o6isi1.