In July, I had the chance to speak at the 2022 ITE International Annual Meeting as part of a session on improving the pedestrian experience at signalized intersections. This topic is so important as for many years, transportation engineers have devalued pedestrian mobility at the expense of vehicular operations and throughput. As cities both big and small are looking to shift users towards walking instead of driving, increasing the convenience of the pedestrian mode will be imperative.
The National Cooperative Highway Research Program (NCHRP) Research Report 969: Traffic Signal Control Strategies for Pedestrians and Bicyclists outlines some of the best practices for enhancing the pedestrian experience at traffic signals. In my first blog of this series, I talked about how to reduce turning conflicts and improve pedestrian safety. In today’s blog, I want to shift focus and talk about pedestrian mobility – specifically how to reduce pedestrian delay. In doing so, I will share some examples and use cases from our citywide multimodal traffic signal optimization project in DC to illustrate how these treatments can be deployed.
How to Reduce Nonmotorized Delay?
The primary objective of traffic engineers at signalized intersections has generally been to minimize vehicular delay. For better or worse, most people are familiar with the primary performance measure at traffic signals – Level of Service (LOS) – which assigns a letter grade to the intersection based on vehicular delay. Pedestrians have a similar LOS letter grade metric, though it is less commonly used than vehicular LOS, undoubtedly due to both the devaluation of the nonmotorized experience and its unavailability in traffic analysis software. Comparing these two metrics, however, shows that pedestrians have a much lower tolerance for delay than drivers. This makes it much more important to reduce pedestrian and bicyclist delay to achieve desirable mobility, safety, and compliance outcomes for nonmotorized users. Here are some highlights from NCHRP 969.
Reducing Signal Cycle Lengths
This treatment is the most significant for reducing nonmotorized user delay. While longer signal cycle lengths can process more vehicles during congested periods (up to a certain point), they can be detrimental to the pedestrians and bicyclists. Determining the appropriate signal cycle length can be a balancing act depending on the number of total phases, as well as the geometry of the intersection.
For instance, increasing the number of phases to provide safer pedestrian crossings will increase the minimum cycle length, as will wider intersections that require longer pedestrian clearance intervals. In DC, many intersections are complex, multileg, or multi-node clusters due to our unique grid, which makes it challenging to maintain a coordinated system with short cycle lengths. Enter the half-cycle – a variation on providing short cycle lengths within a coordinated network. Larger intersections on major arterials may require slightly longer cycle lengths due to combination of intersection geometry, vehicle demand, and total number of phases. However, at smaller or less critical intersections, cycle lengths are set to exactly half of the adjacent arterial cycle length. This can provide a balance between vehicular and pedestrian mobility by providing more opportunities for pedestrian crossings but still maintaining vehicular coordination.
Pedestrian Recall and Maximizing the Walk Interval
Pedestrian recall is used so that the pedestrian phase is served every cycle. Most users can relate to the inconvenience of arriving slightly late to a signal and missing a chance to cross. One simple way to reduce pedestrian delay and eliminate this phenomenon is simply to provide the pedestrian phase each cycle. There are tradeoffs here with vehicular delay on the conflicting approaches, but NCHRP 969 provides a tool to help determine when pedestrian recall may be appropriate based on pedestrian demand and the concurrent vehicular green duration.
To reduce delay even further, considering pretiming pedestrian phases on recall to extend the duration of the Walk interval. Without applying Rest In Walk or a pretiming function, most controllers will default to whatever the minimum programmed Walk value is (typically seven seconds). Extending the Walk interval will increase the percentage of a signal cycle that pedestrians are allowed to cross, in turn reducing their delay. In DC, pedestrian movements with over 0.5 pedestrians per cycle are placed on recall or pretimed, and Rest In Walk is applied to all coordinated phases. As a result, over 90% of signalized intersections do not require a pushbutton activation to cross.
Improve Timings of Multistage Crossings
Multistage crossings are typically found at wide intersections with median refuge islands. While a great tool to reduce pedestrian crashes, signal engineers have traditionally used median refuge islands as an excuse to reduce provided pedestrian crossing time in an effort to minimize vehicular delay. Here are two ways to retime multistage crossings to improve pedestrian and bicyclist mobility.
- Time the Crossing as a Single-Stage Crossing – The simplest way to provide a single-stage crossing through a median refuge is to set signal timings as if the median were not present. The pedestrian clearance interval is timed for a full crossing over the roadway with a low Walk interval. This option does require the longest pedestrian clearances, however and may not be the best balance between the mobility of all modes. Along K Street NW, this approach is used despite the presence of pedestrian signals on median islands.
- Provide a Single-Stage Crossing but Time for Half-Crossings – Where timing for a full crossing would require extremely long pedestrian clearance intervals or in locations with more unique geometry, practitioners can provide a single-stage crossing while timing only for a half crossing. This method provides a Flashing Don’t Walk (FDW) interval for a half-crossing but a longer Walk interval so that pedestrians can get completely through the median before Flashing Don’t Walk begins. Here in DC, we use this approach for crossings of Pennsylvania Avenue (both NW and SE) where median widths range from 15-50 feet wide.
As I stated above, reducing pedestrian delay is going to be critical for increasing the convenience of walking as a mode of transportation. But in order to do that, we have to actually measure pedestrian delay. I encourage you to check out the full report and the latest edition of the Highway Capacity Manual for some methodologies to capture both pedestrian and bicyclist delay. If your jurisdiction does not account for these nonmotorized delay impacts when evaluating projects and alternatives, this document is a great introduction to the concept and can empower you to shift priorities towards multimodal mobility.
Reducing pedestrian delay is an important aspect of the multimodal traffic signal picture – but there’s even more included in NCHRP Report 969. Stay tuned for my last blog in this series where I will focus on how to address bicycle-specific needs!
Read Part 1: Reducing Turning Conflicts For Bikes And Pedestrians
Read Part 3: Addressing Bicycle-Specific Needs – NCHRP Report 969 In Practice