Issues register

EdTech Assessment Toolkit

Bus Routing

Algorithm matching model

In 2017, the Boston Public Schools organised a Transportation Challenge seeking an innovative solution to optimise their bus routes. The Quantum Team from the MIT Operations Research Center won the challenge with an algorithm that utilised Google Maps data and Boston Public Schools data. This algorithm successfully minimised bus routes, reconfigured bus stops, maximised student ridership, and reduced the time empty buses spent on the road (Boston Public Schools, 2017). AI-powered transportation management systems claim to be able to analyse various factors, such as student locations, traffic conditions, road closures, and weather data, to create efficient bus routes and schedules (e.g. AlphaRoute, 2023). These systems dynamically adjust routes in real-time, reducing travel times, minimising fuel consumption, and enhancing overall transportation efficiency.

However, the success of the algorithm led to an unforeseen turn of events. The same system that initially focused on optimising bus routes was later employed to optimise school start times. This decision, approved by the Boston School Committee Office of Equity, entailed significant changes to start times, causing disruptions to family routines, sleep patterns, and before/after school childcare for parents (Ito, 2018).

Snapshot (July 2023)

System task/function: Automate the routing of busses
Model: Algorithm Matching
Deployment: Algorithmic planning of bus routing options to maximise student ridership
Location of application: Boston (US)
Rationale for introduction: Transport students to school safely, reliably, and efficiently
Vendor: Various, e.g. AlphaRoute
Pricing: N/A
Data and computation: Infrastructure data, environmental data, student locations, school operations data
Inequalities/harms: bias (see Ito 2018) for an explanation how this can at times also affect ‘privileged’ groups)
Status: Active
Authority/regulation: District & school level
Unintended consequences: Disruption of family routines, potentially adverse effects such as lower student ridership
Sanction/redress: N/A

References/further reading

AlphaRoute. (2023). Schools. Retrieved from

Boston-Public-Schools. (2017). MIT’s “Quantum Team” Wins First-ever BPS Transportation Challenge with Revolutionary Computer Model. Retrieved from

Ito, J. (2018). What the Boston School Bus Schedule Can Teach Us About AI. Retrieved from