The conclusions are from observing "one route of the Central Ohio Transit Authority (COTA)".
That means it depends strongly on the accuracy of the predictive software used. In principle it could include the chance of, for example, the driver making up lost time.
That means it depends strongly on the accuracy of the predictive software used. In principle it could include the chance of, for example, the driver making up lost time.
The paper is at https://www.sciencedirect.com/science/article/abs/pii/S09658... . It's paywalled.