For the vast majority of people, signal timing is not an exciting subject. Apart from traffic engineers and a select group of superwonks, few people pay attention to or advocate for better signal timing. Whereas the physical designs (and deficiencies) of streets are tangible, signal timing is less discernible. It’s not uncommon for communities to invest significant effort and resources into streetscape projects, only to omit any changes to signal timing.
Yet, signal timing plays a critical role in complete streets. Traffic signals regulate intersections, often the most dangerous conflict points between people driving, walking, and biking. Similar to the role of frequency in facilitating great transit service, signal timing provides (or inhibits) mobility between blocks and neighborhoods. Well-timed signals for bicyclists and pedestrians serve as bridges across busy streets; poorly timed signals isolate people from walking and biking in their neighborhoods, and often lead people to disobey the law and put themselves at greater risk.
In order to discuss what makes a good traffic signal for walking and biking, it is necessary to understand features that determine how traffic signals work:
- Actuated or pretimed: Broadly speaking, traffic signals can be programmed to be pretimed (a standard phasing regardless if vehicles, bicyclists, or pedestrians are present) or actuated (where sensors and call buttons activate a signal phase). Many signals can lie somewhere in between, featuring a shorter timed phase with an extension to accommodate additional vehicles, bicycles, or pedestrians. If a pedestrian phase is built into the signal timing (as is almost always the case in Downtown San Francisco), then the signal is pretimed. If the signal includes a push button for pedestrians or a trigger marking for bicyclists, it usually means the signal is actuated.
- Phase time: The length of time for each phase is typically determined by traffic volumes and whether the phase is actuated or pretimed for pedestrians and bicyclists. Pretimed phases often take longer to accommodate pedestrians: the minimum crossing time is 3.5 feet per second, although many areas around schools, senior centers, and other facilities require 4 feet per second or more. Phases for actuated signals can be as little as a few seconds if only one or two cars are present.
- Cycle length: Cycle length refers to the amount of time a signal takes to run through a full cycle. Traffic volumes, street designs, and actuation (see above) determine cycle lengths. Shorter cycle lengths accommodate all movements more frequently, while longer cycle lengths are usually less desirable and can lead to delays. A 60 second cycle is typical for many intersections in Downtown San Francisco.
- Protected versus permitted turns: The phasing of turns affects conflicts between vehicles, pedestrians, and bicyclists. If turns are protected, then they have a separate phase (a left or right turn arrow) that separates turning traffic from through traffic. If turns are permitted, they may occur simultaneously with through traffic, including pedestrians and bicyclists. Protected phasing result in longer cycle times compared to permitted turns; they are often activated to lengthen the cycle if appropriate.
What does all this mean for livable streets? Pretimed signals with short cycle lengths are typically best to support walking and biking, particularly in urban areas. Protected turns are helpful if a large volume of vehicle-pedestrian or vehicle-bicycle conflicts exist; for very high volumes, scrambles (simultaneous walk and/or bike signals) are appropriate (although these can lengthen cycle lengths). Actuated signals can work well in suburban settings and for low-volume streets intersecting with busier streets. But problems arise when actuated signals are poorly applied in urban areas where streets that should be pedestrian and bicycle friendly are instead designed to prioritize driving. In these circumstances, pedestrian actuation signals are often maligned as “beg buttons” because the ability to walk across the street is effectively the lowest priority.
Unfortunately, Oakland serves as a prime example of how bad pedestrian signal timing can inhibit walkability. Although Oakland has a number of vibrant, dense, pedestrian-oriented neighborhoods and high rates of walking, biking, and transit use, most of its streets still prioritize moving cars fast. In particular, this trait rings true for signal timing. With a select few exceptions Downtown along Broadway and 20th, nearly every traffic signal in Oakland is actuated and requires pedestrians to push a “beg” button to cross the street – including major commercial and transit corridors along Telegraph, Piedmont, Grand, College, and International. This configuration serves as a stark contrast to San Francisco, where nearly all signals are pretimed to always include walk phases.
The intersection of Grand and Bellevue/Parkview Terrace serves as a particularly striking example of what happens when signal timing goes wrong. Located in Adams Point (one of Oakland’s densest districts) adjacent to Children’s Fairyland and Lake Merritt and relatively near BART, the intersection experiences a significant volume of pedestrian activity throughout the day. However, like the car-oriented street design, the signal timing seemingly assumes the opposite. Most traffic occurs on Grand, so the signal seems to default to green for Grand unless otherwise actuated. Yet, perplexingly, pedestrian crossings default to don’t walk, creating a discontinuous pedestrian environment that requires signal actuation every time while walking on Grand (as is the case for many signals on the corridor). To cross Grand, the wait times are extremely long: the cycle length can be greater than 120 seconds and won’t include a pedestrian phase unless pushed with enough time in advance. Furthermore, the intersection only features three crosswalks (instead of four) so it’s possible that crossing the street could take upwards of four minutes, whereas a direct “jaywalking” path would take less than 30 seconds. Remember: this is not a suburban intersection with low pedestrian volumes: it’s the middle of a dense neighborhood adjacent to two of Oakland’s biggest attractions, Children’s Fairyland and Lake Merritt.
The extreme case of the Grand & Bellevue/Parkview Terrace intersection brings us to our main point: when signals are not timed properly, they encourage people to disobey the law and put themselves at risk. Since the signals to cross Bellevue and Parkview Terrace default to “Don’t’ Walk” (despite an obviously long green phase), most people seem to ignore the pedestrian signal and cross regardless – compromising the integrity of the signal and increasing their susceptibility to becoming stranded in the intersection if the signal changes or come into conflict with an unsuspecting turning driver. For those who do follow the law, it unnecessarily penalizes people walking, contrary to the intent of Oakland’s complete streets policy. It’s also worth noting that Oakland is by no means the only example: Berkeley, San Jose, Los Angeles, and many other cities experience a mismatch between pedestrian volumes and signal design.
The good news is that the cost of retiming a traffic signal is a relatively low to moderate in comparison to a full-scale street redesign (the costs vary depending on the scope of timing changes and condition of signal). Most importantly, it requires leadership that is willing to rethink how our streets function and tackle an issue that is incredibly mundane for the vast majority of people. As more cities embrace the goals of Vision Zero, it is likely that signal timing will become more of a focus.