News / Toronto

Five ways cities are tackling 'bus bunching'

You know that annoying problem when three buses arrive at once? It's really hard to solve.

A group of women had to wait longer for the bus on Sherbourne St. in Toronto in January 2017.

Torstar News Service Order this photo

A group of women had to wait longer for the bus on Sherbourne St. in Toronto in January 2017.

We’ve all been there. The bus is late and then three come all bunched together. Internally, we scream: “Why can’t they just keep to the schedule? Isn’t this their job?” As it turns out, bus bunching is an incredibly complicated problem that transit agencies around the world spend millions to solve. There are no easy answers, but here are five promising ways to fix it:

1. Straight Talk: Chicago spent $9 million in 2015 to upgrade the technology that links drivers to the transit commission’s control centre. The city’s 1,800 buses were already equipped with GPS to track location and progress, but now the drivers could speak directly to the command centre if they spotted any issues that would throw a wrench in service.

Outcome: In the first three months of the trial, there was a 40 per cent reduction in big gaps.

2. Analyze That: Miami teamed up with IBM to crunch real-time data in the hopes of spotting bottlenecks and other issues. The system tackled nearly 20,000 bus runs on four major routes, looking at bus paths, speed and stops to assess the accuracy of schedules and bunching alerts.

Outcome: The system was able to predict when bunching might occur up to 60 minutes ahead of the problem, so buses could be redirected accordingly.

3. Bring in the vets: In 2014, Toronto’s transit agency deployed spare vehicles on two of its busiest routes, the 29 Dufferin and 512 St. Clair.

Outcome: Short-turns on the Dufferin bus went from 300 per week to 30 or 40. 

4. Throw out the schedule: Researchers at Georgia Tech told bus drivers to abandon their schedules and just drive with the flow of traffic. In the experiment, buses waited at selected stops or “control points” based on calculations made once the bus arrived there.

Outcome: Gaps between buses self-corrected and the schedule re-equalized without complex intervention from a control centre or pressure on drivers. Drivers also preferred the method, saying they could focus on safely navigating the streets rather than freaking about timing.

5. Customer unservice: Some researchers have tinkered with the idea of imposing boarding limits, or telling passengers they can’t get on the bus even when it’s not full in the interest of getting away from the stop quicker. The researchers who have toyed with that compile papers that look like the chalkboards in A Beautiful Mind. Lots of smart-looking math equations. But try telling someone that an algorithm says they can’t get on the half-empty bus when it’s -40 C.

Outcome: Metro could not find a city that has widely implemented this policy.

More on