A new Slop University paper scores a bike-share operator's depots on a five-level demand-sensing maturity model, finding a depot's score does not predict how quickly its chronically empty racks receive a bike.
A depot-by-depot bike-share audit ties demand-sensing maturity to nothing about how fast an empty rack gets fixed
An overnight bike-share rebalancing algorithm only knows about a rack once someone drops a bike there. A new Slop University paper asks what happens to the racks that never get that first drop, and whether an operator that’s better at noticing the gap is any faster at closing it.
A maturity score is supposed to answer “how prepared is this depot.” Ours turned out to raise a second question we hadn’t budgeted time to ask: prepared for what, exactly.
— Dr Anouk Mensah, Postdoctoral Fellow, School of Emergent Priorities
The University counts the study among the anticipatory-capability agenda’s more sobering instalments: a five-level scorecard that measures a depot’s attentiveness with real precision, and its usefulness to a rider standing at an empty rack with rather less.
Working with School of Emergent Priorities colleagues Dr Lindiwe Achterberg and Dr Iben Chikere, Dr Mensah’s team scored all seventeen depots of a city bike-share operator’s network against a five-level rubric, running from a depot that keeps no record beyond the rebalancing algorithm’s own cache through to one that runs a scheduled reconciliation pass flagging every rack the algorithm has missed. Across a ten-week audit, the team then timed how long a chronically empty rack waited for its first bike once flagged, and found a depot’s place on the rubric explained almost none of the difference.
Dr Iben Chikere said the finding sat comfortably alongside the School’s wider interest in how well an institution can describe a problem without that description doing anything, on its own, to move it.
The full paper is available from the University’s research repository under an open licence, doi:10.5555/slop.zc3a0u.