September 12, 2019
The transit industry is full of performance metrics. On-time performance, miles between accidents, passengers per hour, vehicle miles traveled, level of service — you name it and there’s a metric for it. But for an industry full of stats, an obvious one is missing: headway reliability. There’s currently no standard in how most transit agencies measure the reliability of their headways, and it’s hurting the rider experience.
Riders and experts alike consistently rank unreliable headways and long perceived wait times as leading causes for declining ridership. But what exactly are unreliable headways and long perceived wait times? And how do they creep into an agency’s service in the first place?
It’s time for agencies to ask these questions in earnest and to structure policies that treat poor headway performance according to what’s causing it, mapping specific solutions to each underlying issue. Otherwise, agencies will continue treating the symptoms rather than the root causes of unreliability.
Functionally, unreliable headways can be parsed into two major categories: bunching and gapping. Yet bunched and gapped service are often grouped together under the same umbrella, even though rider perception of one is far worse than the other.
Bunching is annoying, sure, but it’s not the end of the world in itself. It’s a visual reminder of underperforming service reliability and emblematic of operational dysfunction, but riders do get to choose from multiple vehicles. Do you want to get to your destination slightly sooner, or do you want to stretch your legs in the second vehicle with no one else on it? More choice isn’t necessarily a bad thing.
But gapping is entirely negative from the rider’s perspective. Gapped vehicles aren’t as visually striking as bunched buses on top of each other at a stop, but gapped service can mean missing an interview, being late to a doctor’s appointment, or being late picking up a child at daycare. Many passengers even build extra time into their schedules knowing that a 15-minute headway might actually be 20 or 30 minutes. Other riders give up on transit altogether in frustration and look for a different mode of travel to make that trip.
That’s why it’s important to look at bunching and gapping separately, devise strategies for each individually, and prioritize what changes have the biggest actual impact on the rider experience.
Our proposed framework is to have two separate policies that measure bunching and gapping differently. A “bunched intervention policy” could show where bunching is the most severe, showing show service planners and capital planners where to implement bus prioritization interventions, such as queue jump lanes, to speed up downtown travel, preventing vehicles from bunching up too much in denser service environments.
A second “gapped intervention policy” could identify where gaps in service are severe enough to deter riders, which can inform a whole separate set of solutions like short-turning, leap-frogging or other transit jargon not fit for the outside world to proactively smooth out gaps.
Just as on-time performance definitions are structured to put more emphasis on early departures rather than on lates, this would put greater scrutiny on the gaps in service that more negatively impact riders and perceived wait times.
There are so many strategies from a planning, scheduling, and operations perspective to maintain consistent headways that policies need to be structured to highlight where and when to implement them individually rather than only measuring the variability of arrivals at a stop.
What’s amazing is that the data now exists to write, test, and modify these policies today!
A bunched intervention policy should highlight stops on high frequency routes where bunching is approaching unhealthy levels; a gapped intervention policy should isolate stops where gapping is severe enough to start deterring riders. Here are the keys to crafting bunched and gapped intervention policies at your agency:
There’s a lot of data that exists between timepoints that may be lost in analyses. By looking at all stops instead of just timepoints, you can see where bunching starts. This might be a particular turn or intersection that would otherwise be obscured by a timepoint-only analysis.
Define bunched as stops where vehicles are arriving 20% of the scheduled headway. For routes with 10-minute frequency, that means vehicles are arriving at stops every 2 minutes or less. This can not only be frustrating for passengers but can also leave huge gaps behind. Similarly, define gapped as stops where vehicles are arriving at 150% of the scheduled headway. So for routes with 10-minute frequency, that means passengers are waiting at least 15 minutes at these stops. At that point, the service may start to deter passengers. By setting a stricter gapped threshold, you’ll be focusing more on negative rider perceptions.
In order for bunching and gapping intervention policies to work in practice, operations managers, service planners, and long range planners must coordinate on the solutions in the plans. When stops cross the “bunched” threshold, there are opportunities to short-turn the back vehicle or express the front vehicle to mitigate the bunching. Planners should also take note of the stops where that level of bunching starts to spike to supplement bus-priority interventions studies. It is critical to identify these stops where bunching starts to get worse.
Once gapped, routes seldom recover over the remainder of the trip. Without these specific interventions at these locations, the problem would get worse as congestion increases and efforts are wasted at low-priority intersections or stops. When stops cross the Gapped threshold, that would be the intersection to keep a flex (spare) bus to put into service to avoid those delays or look upstream at the road network to find the story behind this gap. This is also an opportunity to coordinate with other micro-mobility providers (think scooters, bike share, etc.) so some passengers are still able to make that trip while longer term fixes are analyzed.
As you reduce bunching and gapping at the most egregious stops, revisit your policies so you can make continuous improvements. The data exists to test different strategies more frequently so take advantage of it and iterate!
Whether agencies are implementing headway-managed service for the first time or just starting to measure where bunching and gapping is most problematic, having real-time and historical bunching and gapping data at the stop level is essential for setting proactive data-driven policies. Such policies enable agencies to zero in on bottlenecks and work towards tangible goals, enabling them to provide passengers with a better, more reliable experience.
For an industry full of stats, metrics and policies, let’s make sure the next ones don’t just supply a letter grade but provide actionable insights needed to build consensus and improve the service.
See how Swiftly can improve service reliability at your agency with an in-depth demo.