A number of public transportation agencies have been experimenting with micro-transit, where buses (or sometimes smaller vehicles) operate on-demand and are dispatched by algorithms. I like the idea – it seems like a possible way to provide service in low-density suburbs, unless we are going to start building differently. However, this op-ed from WHYY says it hasn’t gone well. Keep in mind the author is an advocate for transit riders and transit unions. It’s possible the person is cherry-picking examples or that the pilots in question for poorly implemented and managed.

On average, microtransit pilots across the U.S. have a ridership of zero to three riders per hour, with most pilots operating much closer to zero than three. For comparison, the Route 127, one of the most confusing and infrequent buses in SEPTA’s network, still moves an average of 13.9 passengers per revenue hour. When AC Transit in Oakland, Cal. replaced one of its  low-performing fixed-routes with microtransit, the per passenger subsidy more than doubled. And when Kansas City attempted microtransit, the ridership was so low that by the end of the pilot, they ended up paying $1,000 per passenger to operate the service.

WHYY

To be fair, this article is specifically arguing against implementation of this option by SEPTA (Southeast Pennsylvania Transportation Agency), which is not known for above-average implementation or management. The currently have an app which provides real-time data on bus and train arrival, but the data seems to be supplied by a random number generator. So I would not be too hopeful that they would be the first to pull this off successfully. Maybe they should just give everybody a pre-paid card to use Uber, or hire small-time taxi drivers who lost their life savings when that industry was upended a few years ago.