Transit Network Design Course Highlights

February 7th, 2013

On January 17th and 18th, two staffers from Metro’s Office of Planning attended a two day transit network design course, offered by long time transit planning consultant, transit blogger and now author, Jarrett Walker.  In transit planning circles, Walker’s recent efforts, culminating with “Human Transit,” have been very well regarded.  For many planners, his book has been a breath of fresh air in helping to demystify how complex transit offerings can be made more simple, customer focused, effective, and useful for everyday city life.  Many planners have an appreciation of the attention he has given to linguistics, and how word choice (i.e., the use of “transfer” vs. “connection” or “transit route” vs. “transit line”) can subtly reinforce or undermine certain collective beliefs about the usefulness of transit, or anything else for that matter.

The participants in the DC course were fairly diverse, although all of those in attendance had a vested interest in transit in some way.  Among these were urban planners, consultants, advocates, and transit planners, among others.  Care was taken to ensure that each working group had a included at least one transit professional mixed with other disciplines and backgrounds in order to facilitate a balanced discussion. Mr. Walker covered a number of topics including:

  • the relationship between density and transit demand
  • the trade-off between frequency and coverage
  • the link between frequency and transit’s usefulness, particularly in urban areas less reliant on cars
  • why (particularly fixed guideway) transit service to airports is often seen as a regional priority
  • anchoring transit lines with important destinations on both ends to boost transit efficiency
  • the use of “pulse points” to reduce transfer penalties between infrequent bus routes
  • the concept of transit “operating efficiency”, or getting the most bang out of your transit buck
  • peak period vs. all day service
  • the difference in transit rider perception of wait time and in-vehicle time
  • why capital projects generate the most attention but why operations are the principal drivers of transit budgets
  • how branching divides service frequency, and other facts of transit geometry
  • the importance of regional context and orientation of useful transit service for TOD planning

The first challenge presented to us was to design a transit network for a mid-sized city of about 400,000 people given a very modest operating budget (represented by 25 buses).  The fairly limited resources provided for this first task presented some real challenges for each of the groups.  Should we strive to offer a dense network of frequent service in the dense urban core (filled with its central employment hub, dense housing, university, and hospital) to maximize ridership, or do we prioritize coverage of the more suburban areas as a goal for this hypothetical city?  In a world of limited resources, maximizing both ridership and coverage are not possible, and this initial exercise forced each group to explicitly address this challenge head on.  Although every transit agency must address the ridership vs. coverage issue, Mr. Walker noted that very few actually tackle it explicitly.  In other words, few transit boards are asked to direct their agencies to specifically allocate a portion of their annual operating dollars to the coverage goal, with the rest being allocated to high ridership services typically found in and near the urban core.

Exercise 1: Allocate 25 buses

The initial and subsequent interactive exercises introduced other concepts and questions to consider.  Should we introduce a pulse system as a means to connect mostly infrequent bus services with timed connections to reduce passenger wait times?  If so, where does it make the most sense to place this transit center with its variety of transit services?  What about the concept of the frequent grid, something that Mr. Walker has often highlighted?  If we employ a frequent grid, what is the optimal spacing for these transit lines?  What frequency should those lines have?  Is there a single network design solution that solves multiple problems and create multiple connections?  What is the role of limited stop services (similar to the Metro Extra brand)?  Where should we allocated fixed guideways for high capacity, rapid transit services (similar to Metrorail)?  For a medium-sized city with a $1.5B capital budget, should these be busways or light rail?  Should bus transitways be open to local buses or restricted only to BRT buses?  On the topic of frequency, one memorable moment included Mr. Walker’s rather emphatic argument that the frequency of a transit line was an existential matter, rather than one service characteristic among many others.  He made the case that the more frequent a transit line, the more it exists in space.  Alternatively, a very infrequent transit line hardly exists at all, save maybe once per hour.  Anyone who relies on transit to get around understands that frequency is important, but I had not heard it put quite such an elemental way before.

Taking the course was a challenging, fun, and rewarding experience.  It’s highly recommended it to anyone who has a role or strong interest in transit, be they a transit planner, urban designer or architect (especially them!).  We will be using the skills gained from this training in a variety of projects, including the Metrobus Market Effectiveness Study and upcoming bus/BRT corridor planning work.


Related Posts:

Categories: In The News Tags: , ,
  1. Alex
    February 10th, 2013 at 00:10 | #1

    I appreciate the comment about frequency being existential. This is far and away the most frustrating thing about Metro. If the trains run more frequently, I would bet that the concern people express about all the other problems would diminish.

  2. Steve Strauss
    February 12th, 2013 at 11:10 | #2

    Any discussion of WMATA’s frequent use of loops and meandering routing of buses? It gets to the coverage vs. frequency issue.

  3. Steve Strauss
    February 12th, 2013 at 11:12 | #3

    It would be interesting to know which lines Alex rides and describes as having infrequent Metrorail service.

  4. Jonathan
    February 12th, 2013 at 13:01 | #4

    Alex, are you thinking of a particular Metrorail or Metrobus line that does not run frequent enough? With respect to both rail and bus, we hope riders will take a look at our new (December 2012) regional bus maps posted online, which help indicate the corridors where buses are more frequent and tend to be closer to ‘schedule-free’ service. In particular the “framework service” map on the back of each jurisdictional map, presents a good picture of places one can travel in the region on reasonably frequent, all day buses and trains.

    Steve, we didn’t discuss Metrobus service as part of the course work, rather a fictional city called “Newport.” Mr. Walker noted that his course tends to be more useful discussing a fictional city because it avoids preconceived notions of how existing networks are designed, and focuses attention on the concepts noted above. The topic of providing service on DC’s L’Enfant-penned diagonal arterial streets did come up, but only in an informal discussion. As far the Metrobus network structure is concerned, in some cases the street network restricts the route structure, while in others there may be inefficiencies in the network that have developed through service and/or routing changes over the years. However, the framework service map does bring forward a coherent network structure of the region’s major bus and rail lines that perhaps wasn’t previously apparent to the casual observer. We hope to explore some of the course concepts in the Metrobus Market Effectiveness Study, now underway.

  5. February 28th, 2013 at 18:37 | #5

    Thanks for paying attention to structuring the bus network — Jarrett Walker’s great ideas about making bus networks legible are almost the exact opposite of what I’ve found with the impenetrably complex Metrobus network. I sometimes think each bus is an enigma: I have almost no way of knowing where, when, or how any given bus will go, and therefore avoid ever boarding in order to prevent unpleasant surprises. (That kind of defeats the purpose of a transit network.) Even the few attempts to make the service patterns more legible, like the “spider map” at my local Metrorail station (Waterfront), have ended in frustration when I found that the most useful route shown on the map only runs once, only on weekday mornings. It’s no wonder that people flock to Metrorail and Circulator services, even for trips that Metrobus could serve quite effectively.

    However, I must give huge kudos for the new and much easier to read bus maps. Is there a chance that we could get the Framework Service map printed separately? I loved exploring LA with just the “Every 15 Minutes (or Less) Map” in hand.

  6. Jonathan
    March 1st, 2013 at 12:45 | #6

    Payton, thanks for your comment. You make some good points on the legibility of the transit network and more to the point, the bus system. The system should be easy to use and understand and we’re working to make that a reality. You note the new bus maps, which we think help riders understand the usefulness of the bus network in a way that hasn’t been available before. It may be helpful to either print out one of those maps or perhaps keep a link on your smartphone (if you have one). Each version of the bus maps has the “framework service” map on the 2nd page or back, so it’s available through each link. We do this because the maps are primarily designed to be used as hard copies, though we are exploring ways to perhaps create an interactive map on the website, instead of just the having PDF files. I’d also note that you can use the trip planner and get easy access to real time bus arrival information through the WMATA site or a 3rd party application. The combination of these tools can help users navigate the system.

    On a personal note, I’m new to the region, and even though I work for Metro, there have been times when the bus network hasn’t been clear to me at first glance. Where does the 92 terminate? If I take the 7X, can I get back home on that bus? It can be confusing at times, but employing some of the tools available and focusing on the major bus lines in the region (found on the framework service map), the overall network begins to make more sense.

    Still, we can always improve and are working on that. We are investigating how to be able to incorporate the new frequency-based maps at the stop level, to push the information out to the street, so that riders can see the interaction between their stop and the broader system. As I mentioned above, we have begun a study to chart out the future of the Metrobus network. We are also working on a (draft) strategic plan, Momentum, which is proposing several initiatives to improve the system by 2025. One that might interest you is an effort to upgrade communications throughout the system. We’d love to hear your thoughts about this and other Momentum projects at

Comments are closed.