1 of 17

Making Headway:

Exploring Headway Management at the MBTA

Sandy Johnston, Deputy Director of Regional Transit Planning

Arthur Prokosch, Senior Manager of Fare Tariff

Draft for Discussion & Policy Purposes Only

*personal bios and intros*

You might be wondering, how did a couple of people who don’t work in Operations stumble into this topic? For one, both of us are regular bus riders who care about solving the bunching issue. But we came into it through the MBTA’s Lean Six Sigma Green Belt program, which offers select folks from across the agency an opportunity to work on knotty process improvements problems in a cross-functional environment and through an organized philosophy of change management. We were lucky enough to work on a really collaborative and multi-disciplinary team whose makeup really reflected the complexity of this topic. We worked on this in a focused, though part-time, manner for several months in addition to our usual responsibilities, under the sponsorship of Mike Joyce, a supervisor in the Bus Operations Control Center.

2 of 17

Bus Bunching

TCRP Synthesis 155 Pg. 9

2

Draft for Discussion & Policy Purposes Only

I imagine that many people in the audience know what bus bunching is and how it works, but here’s a quick overview. Thanks to Dr. Kari Watkins, who you’ll hear from later, for the graphic. In the graphic, each row shows relative position of buses as they proceed along a route
Bus service is typically advertised as appearing at a certain interval, every X number of minutes, known as a headway.
Bus service on the road, especially service that operates in mixed traffic, can be exposed to a wide variety of factors that can introduce variance into the schedule. Some of these include:

Variance in passenger boarding times
Traffic
Signal timing
Need to load a wheeled mobility device.

Without generous recovery time at the end of a route, these factors tend to reinforce each other and cascade across the schedule, leading in some cases to buses essentially chasing each other’s tails all day long
From the passenger perspective, this looks like bunching—two or more buses arriving in a time span well below the scheduled headway—followed by gapping—a time between buses much longer than the scheduled headway.
Bunching is a highly visible operational challenge for agencies. At the T, we’ve heard about it being a problem time and again from advocates and elected officials—and our own data bears out that it’s a challenge. But it’s also a complex problem to unwind, which we’ll discuss a little bit today.

Points that we would ideally cover in this intro:

Keeping this slide, I'd emphasize that each row shows relative position of buses as they proceed along a route. The red bus represents a small perturbation that positive feedback loops amplify, etc.

�1) Bus speeds are influenced by myriad dynamic factors, thus variation is inevitable.�2) Bunching and gapping are flip sides of the same coin (variability).�3) Feedback loops amplify any variation. The largest is passenger boardings, which accumulate in gaps and then cause boarding/alighting delays. This means variable arrivals cause variable crowding.�4) Without sufficient time (and space) for vehicles to recover at the end of their trips, bunching cascades throughout the day.

A key challenge in the operation of high-frequency bus service is bus bunching, also referred to as bunching and gapping. Bunching is a phenomenon in which adjacent buses tend to form pairs (Newell and Potts 1964; Chapman and Michel 1978).

From the passenger perspective, this pairing effect leads to large gaps in service followed by the arrival of two (or more) buses in short succession. Bunching often starts with a small perturbation, such as a group of boarding passengers, which delays a bus.

Figure 2 shows four buses initially spaced at an even headway, h. The second bus then picks up some delay that puts it behind its headway target by δ. Now the second bus has a larger forward headway, h+δ, and the third bus has a smaller forward headway, h-δ. On average, the second bus will now encounter a few extra passengers who might otherwise have boarded the third bus. The extra dwell time associated with these additional passengers will put it further behind.

Similarly, the third bus will encounter fewer passengers and tend to run a little ahead. With no intervention, the second and third buses will eventually converge, opening large gaps in front of and behind the bunch

3 of 17

Critical Finding

We Promise: �Headways

Riders expect “walk-up” service on frequent bus routes
Promised by:

Bus stop signs
PDF schedules
Service Delivery Policy
Transit App

There is a misalignment between our goals for frequent bus service and the way we deliver that bus service

We Deliver: �Schedules

Inspectors and dispatchers manage individual trips to schedule
Tools & staffing metrics�reinforce this practice

3

≠

Draft for Discussion & Policy Purposes Only

If there’s one thing from this presentation I want people to take away, it’s this slide. The challenge that we face at the MBTA is that the type of service we promise to our riders—buses arriving at a regular headway—is not the kind of service that our workforce, tech, and organization systems are built to deliver. We just on December 15 implemented Phase 1 of a systemwide bus network redesign, and as part of that we are promoting newly branded and scheduled “frequent routes” that promise service at least every 15 minutes for the entire span of service. But our staff who manage buses—on-street inspectors and supervisors, and dispatchers in OCC—are using legacy training and tools optimized to support a strict written schedule, and operators are incentivized and managed to measure their performance relative to that schedule. So when things go wrong—and they do, every day—buses inevitably bunch and riders experience gaps. There is significant interest in changing this—but again, it’s a complex issue that touches numerous parts of the organization.

4 of 17

MBTA Progress on Bus Headway Management

We are trying to learn!

4

Asking questions

Institutional learning

Experimentation

Adoption

Scaling systemwide

We are here

Draft for Discussion & Policy Purposes Only

5 of 17

Research Take-homes that Defined Our Pilot

5

Peer Agencies�SFMTA�CTA�Metro Transit (MN)

MBTA Dispatchers and Inspectors are overloaded compared to peer agencies
Technology changes are powerful, but take time

Previous MBTA Pilots�Route 77, August 2024�Route 1, Marathon Day 2024

Dispatchers or Inspectors need clear roles and goals to implement headway management

Literature Review�U.S. & International Experiences

Terminal holds can be the most direct and effective adjustment strategy (followed by short turns and brief mid-line holds)

Plus background information from

Discussions with Bus Operations, Ops Analytics, and Operations, Planning, & Scheduling leadership;
Discussions with operational tech teams; and
Field observations at Nubian Square terminal(5/30/2024)

Draft for Discussion & Policy Purposes Only

To inform our process improvement project, we turned to our peer agencies—some of whom are here on the panel today, so Jonathan and Joseph in particular, thank you for your help—to the research literature, and to observation of our own system for guidance. For the sake of time we won’t go over everything we learned, but all of these lessons directed us toward a few key insights:

Staff capacity, especially for dispatchers, is critical, and we are understaffed for achieving headway management
Staff involved in headway management need clear roles, and ours don’t
Terminal holds—a dispatcher or inspector asking a bus to hold at a route endpoint to achieve adequate spacing—can be a simple, testable, and controllable way to start with headway management.
Technology is a key part of headway management, but not a silver bullet, and takes time to improve and implement.

6 of 17

Maintaining Headway is a Complex Challenge

Root causes inside the circle are being ignored, threatening progress on every other item

6

People

Methods/Policies

Measures

Machines/Equipment

Technology

Environment

Manpower

Lack of training

No clear roles

No procedures for maintaining headways

Unrealistic schedules

Distorted/

untrusted OTP metric

Unpredictable traffic & ridership

Diversions

Dispatching tools do not show headways

Tracking issues / ghost buses

Operator displays do not show headways

Uneven �Headways

Draft for Discussion & Policy Purposes Only

7 of 17

Piloting Headway Management on Route 1

6 AM to 6 PM, Wednesday, September 4, 2024

Pilot Plan

OCC dispatcher dedicated to Route 1 with other responsibilities reduced
Objective: Ignore the schedule and maintain even headways
Dispatcher radioed adjustments directly to Operators or via the Nubian Inspector

Logistics

Bus Ops (including Operators), OPSS, and TID informed of the pilot plan
Project team and other interested MBTA staff recorded observations throughout the day

7

Check for feasibility to hold/deadhead: at/near terminal, trailing headways, few/no passengers

Review situation, revising hold length or repeating for compliance as necessary

Instruct the target bus to hold until HH:MM, depart before schedule, or deadhead to 3^rd timepoint

Start: scan route for bunching/gapping

Place block waivers on any missed service

Draft for Discussion & Policy Purposes Only

Informed by research and peer insight, we decided to conduct a one-day pilot. The goals of the pilot were twofold: first, to test a discrete element of headway management, namely—terminal holds, and second, to observe challenges and successes and learn more about organizational challenges and opportunities. We chose Route 1, which runs between Harvard Square and Nubian Square in Boston’s Roxbury neighborhood. It’s a critical, frequent route that is relatively simple to manage because it has no variants and a relatively linear route. It’s also one of our most bunched routes, and a key crosstown link between various important destinations and rail transfers.

We chose Wednesday, September 4, because it was the first Wednesday of the fall schedule rating, but the day before Boston Public Schools start, so that we could avoid affecting kids if anything went wrong. We essentially empowered dispatchers by telling them to throw out the written schedule and use terminal holds—and other tactics like short-turning, midpoint holds, and expressing, at their discretion—to manage to a headway. OCC dedicated a dispatcher just to this route for the duration of the pilot. We made sure to inform all relevant staff, including operators and field inspectors, of the pilot before it commenced. Our project team and other T staff observed the pilot in the field and from OCC.

8 of 17

Pilot Results

Headway improved and we learned more about the challenges to implementing headway management on a larger scale

8

*This analysis defines 'Bunched' as less than 25% of a scheduled headway, and 'Gapped' as over 175% of a scheduled headway.

AVL data comes from TransitMaster, available on https://mbta-massdot.opendata.arcgis.com/

Quantitative Findings

38%-40% reduction in bunching* compared to Wednesdays after the pilot (shown) or over the year before
90% or more of non-Pilot Wednesdays had more bunching, gapping*, or both

Draft for Discussion & Policy Purposes Only

9 of 17

Pilot Results

Headway improved and we learned more about the challenges to implementing headway management on a larger scale

9

Qualitative Findings

Dispatcher attention and capacity is a scarce resource

Communication between Dispatchers, Inspectors, and Operators is complicated by a lack of clear policies and responsibilities

Existing technology does not show headways, making adjustments harder to plan and less precise

Most adjustments are trade-offs that inconvenience some riders or personnel

These challenges can compound each other!

Draft for Discussion & Policy Purposes Only

Reflecting the core insight that institutional alignment and people management is as important as number crunching in achieving headway management, we made sure to make some qualitative observations in addition to our data analysis. Many of these observations, from the day of the pilot, mirror our initial observations from the research phase of the project.

(read off slide)

Roughly one-third of adjustments were not implemented exactly as intended
Some riders will wait longer to benefit many others
Lack of layover space at terminals limits terminal holds

Though our data analysis shows that the fundamental question of the pilot—can headway management through terminal holds improve performance—has been answered in the affirmative, these qualitative findings form the core of our ongoing recommendations for the MBTA.

10 of 17

Roadmap to Headway Management

Adopting the lessons of this project and transitioning to headway management is a long project. Here are our recommendations for getting there.

10

(1) Clarify roles and responsibilities

Identify an executive champion

Ensure that managing headways is someone’s job

Train staff appropriately and give them the necessary support and skills

(2) Improve and socialize metrics

Choose & refine dispatching performance metric

Adopt Excess Wait/Trip Time for bus

Continue to explore better metrics for headway management

(3) Build out tech to support headway goals

Develop and enable headway management features in Skate

Understand and enable TransitMaster headway view

Adapt Operator screens (MDTs) to show headway info

Adapt real-time predictions / public screens

(4) Adopt knowledge from other transit agencies

Deepen contacts with peer agencies transitioning to headway management

Participate in inter-agency working group identified during this project

(5) Conduct a longer-term pilot

Test and evaluate different strategies for headway management

Familiarize staff with headway management

Observe tech and organizational needs

(6) Prepare agency for systemwide deployment

Strengthen connections between policy, planning, and Operations

Review procedures for operationalizing new policies and documentation

Ensure that metrics and policies are socialized with all relevant operational departments

Draft for Discussion & Policy Purposes Only

11 of 17

OCC Immediate Actions

Steps currently underway due to project findings

11

(1) Clarify roles and responsibilities

Identify an executive champion

Ensure that managing headways is someone’s job

Train staff appropriately and give them the necessary support and skills

Building Consensus throughout all relevant departments

For winter rating, OCC is implementing specific headway management responsibilities for dispatchers

Revising OCC SOP 92 to clarify headway management strategies

Establishing KPIs to measure and improve dispatcher effectiveness

OCC is actively hiring with a goal of 26 active dispatchers

Two dispatchers currently in training

Goal of increasing dispatcher count on all shifts, with a dedicated BNR headway manager

Current Status

Draft for Discussion & Policy Purposes Only

As I mentioned before, our project sponsor was a supervisor in Bus OCC; we chose to concentrate power in the hands of dispatchers for our pilot; and we emphasized the need for staffing up in OCC as an early action item in adopting headway management. I’m happy to report that our Ops department is moving forward with a number of positive actions in OCC to lay the groundwork for headway management, including:

Hiring, I believe, 7 more dispatchers
Revising existing SOPs and developing new ones
Establishing metrics to evaluate dispatchers, for the first time ever
Designating one dispatcher on all shifts to manage headways on frequent-service routes, something the T used to do (on what were then called Key Bus Routes) but had discontinued

But there’s a lot to do across the rest of the organization as well.

We’re looking forward to seeing how this project continues to move forward—and to hearing from all of our peers here today.

12 of 17

Thank You!��Contact:�Sandy Johnston sjohnston2@mbta.com�Arthur Prokosch aprokosch@mbta.com

Draft for Discussion & Policy Purposes Only

13 of 17

Appendix: Extra Slides/Info

Draft for Discussion & Policy Purposes Only

14 of 17

MBTA “Making Headway” Project Team

14

Tracy DiPetrillo, Acting Chief of Staff (OCAO)

Sandy Johnston, Deputy Director of Regional Transit Planning (PSP)

Michael Joyce, Division Chief of Operations Control Center (OCC)

Sean Kurnas, Business Analyst (TFM)

Arthur Prokosch, Senior Manager of Fare Tariff (PSP)

MBTA Lean Six Sigma Green Belt Program

Selective program
Offered multiple times per year
Allows MBTA staff to develop change management skills and work intensively on a project outside their usual role

OCAO = Office of the Chief Administrative Officer PSP = Policy and Strategic Planning

TFM = Transit Facilities Maintenance OCC = (Bus) Operations Control Center

Draft for Discussion & Policy Purposes Only

You might be wondering, how did a couple of people who don’t work in Operations stumble into this topic? For one, both of us are regular bus riders who care about solving the bunching issue. But we came into it through the MBTA’s Lean Six Sigma Green Belt program, which offers select folks from across the agency an opportunity to work on knotty process improvements problems in a cross-functional environment and through an organized philosophy of change management. We were lucky enough to work on a really collaborative and multi-disciplinary team whose makeup really reflected the complexity of this topic. We worked on this in a focused, though part-time, manner for several months in addition to our usual responsibilities, under the sponsorship of Mike Joyce, a supervisor in the Bus Operations Control Center.

15 of 17

Critical components of frequent and reliable service

Vehicles

Dispatching

Staffing

Levels

Travel

Times

Sufficient dispatchers, technology, and policies to manage day-to-day service issues

Dispatching

The number of vehicles available and ready for service

Vehicle Availability

Sufficient frontline management & operators to run scheduled service

Operators &

Front Line Management

The time it takes for a trip to make it from end-to-end. For rail, major drivers are state of tracks, signal, and power infrastructure. For bus, major driver is traffic and transit priority.

Scheduled vs. Actual Run Times

15

Draft for Discussion & Policy Purposes Only

16 of 17

The Challenge: Uneven Headways Hurt Service Reliability

Bunching and gapping are major contributors to an uneven, often poor rider experience, attracting negative publicity

16

Customers Experience

Bus bunching

Long wait times

Crowding

On a typical day at the MBTA, buses on the same route leave terminals less than 1 minute apart

130-190 times

Main suspects include:

1
111
57
SL5

https://commonwealthbeacon.org/opinion/how-to-put-an-end-to-bus-bunching/

https://transitmatters.org/blog/reveal-mbtas-slowest-most-bunched-bus

“I often get complaints about bunching and related gaps in service on MBTA buses… Bunching is a perennial challenge for MBTA bus operations”

https://willbrownsberger.com/bus-bunching/

Draft for Discussion & Policy Purposes Only

17 of 17

Benefits and Risks

Improving headways is critical to performance, ridership, and the strategic goals of the MBTA

17

Better customer experience

More efficient operations

Positive perception of MBTA

Continued poor bus performance

Lower ridership

Strategic risk to Bus Network Redesign

Benefits

Risks

IF WE DON’T

IMPROVE HEADWAYS

IF WE DO

IMPROVE HEADWAYS

Draft for Discussion & Policy Purposes Only