Skip to main content
Roadside speed signs stop working after eleven weeks

Roadside speed signs stop working after eleven weeks

A new Slop University paper estimates the effect of driver-feedback speed signs across 61 residential streets, finding a 2.41 km/h reduction in the first fortnight and no measurable effect from week eleven, against a fleet median deployment of 11.6 weeks.

A study of 61 residential streets finds the effect gone by week eleven, and a trailer roster that moves each sign at 11.6

Every driver-feedback sign in one metropolitan council’s fleet is relocated at a median of 11.6 weeks. A paper published today by the School of Continuous Improvement reports that the effect it was relocated to deliver has been over for a fortnight by the time the trailer moves.

The study follows the council’s six sign trailers across 61 residential streets over three years, using permanent traffic counters that were installed before the program began and were never available to the depot that scheduled it. Because streets were reached in the order the trailers came free rather than in any order of need, the researchers were able to compare each street against the streets still waiting. Mean speed fell by 2.41 km/h in the fortnight after a sign arrived, then recovered steadily, and from week eleven the study can no longer measure any difference at all.

The interesting part of this is not statistical. Two numbers were arrived at by people who never met, for reasons that had nothing to do with each other, and they landed a fortnight apart.

— Dr Torun Ezeigwe, Senior Lecturer and Director, Master of Applied Measurement

The University counts the work among the applied-measurement scholarship it has been consolidating across its schools, and regards the decay estimate as a contribution its partners in local government are well placed to use. Dr Thandiwe Solberg, Senior Lecturer and Convenor, Demo Quarter, noted that the depot’s own records made the analysis possible: “The roster was handed over without anyone asking us what we wanted it for, which is more than most operational documents are offered up for.”

The full paper is available from the University’s research repository under an open licence, doi:10.5555/slop.07lgun.