If you live in a major city, I can take a pretty good guess at one of your most common frustrations: traffic. In city driving, the journey is rarely better than the destination. In most cases, we just want to get where we’re going. Traffic is not just frustrating, but it has consequences to the environment as well. All those idling vehicles have an impact on air quality. When you’re stuck and sitting behind a long line of cars, it’s easy to let your mind wander over solutions to our traffic woes. But, traffic management in dense urban areas is an extremely complex problem with a host of conflicting goals and challenges. One of the most fundamental of those challenges happens at an intersection, where multiple streams of traffic - including vehicles, bikes and pedestrians - need to safely, and with any luck, efficiently, cross each others’ paths. Over the years we’ve developed quite a few ways to manage this challenge of who gets to go and who gets to wait, from simple signs to roundabouts, but one of the most common ways we control the right-of-way at intersections is the traffic signal.
There are a lot of good analogies between cities and human anatomy, and roadways are no exception. Highways are like the aorta with a high capacity and single major destination. Small collector roads are like the capillaries with not much capacity but a connection to every individual house and business. And, in between are the aptly-named arterial roadways, the medium-capacity connections between urban centers. Rather than ramps, overpasses, and access roads to control the flow of traffic, arterial roads use at-grade intersections through which only a few traffic streams can pass at a time. We call this “interrupted traffic flow” for obvious reasons. In most cases, these intersections are the limit to the maximum throughput of the roadway. In other words, increasing the number of lanes or the speed limit won’t have any effect on the overall capacity of the road. The only way to increase the number of vehicles that safely travel from point A to B is to increase the efficiency of the intersection. In addition, these intersections are where a vast majority of accidents occur. For these reasons, traffic engineers put a lot of thought and analysis into the design of intersections and how to make them as safe and efficient as possible.
Controlling the flow of traffic through an intersection, otherwise known as assigning right-of-way is an enormous challenge and almost always requires a compromise of numerous conflicting considerations, including space, cost, approach speed, cycle time, sight distance, types and volumes of traffic and human factors like habits, expectations, and reaction times. Intersections also need to be rigidly standardized so that, when you come to an unfamiliar one, you already know your role in the careful and chaotic dance of vehicles and pedestrians. From a throughput standpoint, the ideal intersection would cause no interruption in flow whatsoever, but you can’t put a high-five interchange on every city block. On the other hand, simple signs are cost-effective and don’t require any extra space, but they can’t handle a lot of volume because they create an interruption for every single vehicle passing through the intersection.
You can see why traffic signals are so popular. They aren’t a panacea for all traffic problems, but they do offer a very nice balance of the considerations we discussed before: Relatively low cost, minimal space requirements, and able to handle large volumes of traffic with only some interruption. In their simplest form, traffic signals are a set of three lights facing each lane of an intersection. When the light is green, that lane has the right-of-way to cross. When the light is red, they don’t. The amber light warns that the signal is about to change from green to red. Beyond this basic function, traffic signals can take on innumerable complexities to accommodate all kinds of situations. Let’s take a look at a typical intersection here in the U.S. to show how they work.
At each approach to the intersection, there are three directions vehicles can go called movements: right, through, or left. Right and through are usually grouped together as a single movement, so a typical four-way intersection has 8 vehicle and 4 pedestrian movements. These movements can be grouped into phases of the traffic signal. For example, the left turn movements on opposite approaches can be grouped into a single phase because they can both go at the same time without conflicts. Traffic engineers use a ring-and-barrier diagram to sketch out how different phases of the signal are allowed to operate. Here’s a ring-and-barrier diagram for our example intersection. The first phase is the major street left turns, then the major street vehicle and pedestrian through movements, a “barrier” to clear the intersection, the minor street left turns, the minor street vehicle and pedestrian through movements, and finally another “barrier” before the cycle starts again. There are an endless variety of phasing arrangements that traffic engineers use to accommodate various intersection configurations and traffic volumes for each movement. Even the simple decision of whether to use protected or unprotected left turns takes a significant amount of analysis and consideration.
Another important decision is how long each sequence of a phase should last. Ideally, a green light should last at least long enough to clear the queue that built up during the red light. This isn’t always possible, especially during peak times on busy intersections. In these cases where the intersection is saturated, the green light might be extended for each phase to minimize the startup and clearance times, which are periods when the intersection isn’t being utilized to its maximum capacity. The amber light needs to last long enough for a driver to perceive the warning and decelerate their vehicle to a stop at a comfortable rate. One second for every 10 miles per hour or 16 kilometers per hour on the speed limit is a general rule of thumb, but traffic engineers also take into account the slope of the approach and other local considerations when setting the timing for yellow lights. In most places in North America, you are allowed to enter an intersection for the full duration of a yellow light, which means there needs to be a time when all phases have a red light to allow the intersection to clear. This clearance interval is usually about a second but can be adjusted up or down based on speed limit and intersection size.
So far we’ve only been talking about signals on a set timing sequence, but most traffic signals these days are more sophisticated than that. Actuated signal control is the term we use for signals that can receive input from the outside and use that information to make decisions about light timing and sequence on the fly. These types of signals rely on data from traffic detection systems. These detectors can be video cameras or radars, but most commonly they are inductive loop sensors embedded into the road surface. These are essentially large metal detectors which simply measure whether or not a car or truck is present, sometimes to the annoyance of bicycles, scooters, and motorcycles that may be too small to trigger the loop. Whatever the type of sensor, they all feed data into an equipment cabinet located nearby. You’ve probably seen hundreds of these cabinets without realizing their purpose.
Inside this cabinet is a traffic signal controller, essentially a simple computer that is programmed with specific logic to determine when and how long each light will last based on the information from the detectors. Actuated control gives a traffic signal much more flexibility to handle variations in traffic load. For example, if a nearby road is closed and traffic rerouted through a signal that doesn’t normally see such a high demand, it may need to be reprogrammed before the closure. A light equipped with actuated control will simply see the additional traffic and adjust its phasing accordingly. Same thing with special events, like concerts and sport games, that create huge traffic demands on irregular schedules, and even seasonal changes in traffic, like in major tourist destinations. Actuated systems can also keep you from waiting at a long light when no one’s crossing in the other direction. Finally, actuated control can help by giving priority to emergency vehicles and public transportation by using specialized detectors, like infrared or acoustic sensors, that communicate directly with certain types of vehicles.
But, actuated control isn’t the end of the complexity. After all, it still treats each intersection as an isolated entity, when in reality each signal is a component of a larger traffic network. And each component of the traffic network can have impact, sometime a major impact, on other components in the system. Take the classic example of two signals closely spaced in a row on a major roadway. If one signal gives a green but the next one doesn’t, cars can back up. If they back up far enough, they can sit through multiple cycles at an intersection without being able to pass through until the light beyond clears. It’s a frustrating experience for anyone: a signal is inadvertently, but significantly, reducing the capacity of an adjacent signal. One solution to this problem is signal coordination where lights can not only consider the traffic waiting at their intersection but also the status of nearby signals. This is a very common configuration on long corridors with relatively minor, but frequent cross streets. The signals on the major road are timed so that a large group of vehicles, called a platoon by traffic engineers, can make it all the way through the corridor without interruption. This type of signal coordination can significantly increase the volume of traffic that can pass through intersections, but it really only works on stretches of road that don’t have a other sources of traffic interruptions like driveways and businesses. If the platoon can’t stick together, the benefits of coordinating signals mostly get lost.
The obvious next step in efficiency is coordination of most or all the signals within a traffic network. This is the job of adaptive signal control technologies, or ASCT. In adaptive systems, rather than individual groups of lights, all the information from detectors is fed into a centralized system that can use advanced algorithms, like machine learning, to optimize traffic flow throughout the city. These types of systems can dramatically reduce congestion, but they’re only just starting to be implemented in major urban areas. As sensors become more ubiquitous and computing power increases, traffic management may slowly but surely be relegated from civil engineers to software developers and data scientists. But, that also means that ASCT systems may be more vulnerable to security threats, a scary thought if they’re controlling the signals for an entire city.
On the complete opposite side of centralization, many believe that self-driving cars are the next revolution in traffic management. If every vehicle could communicate and coordinate with every other vehicle on the road, interrupted traffic control could eventually become a thing of the past. But don’t get your hopes too high. In dense urban areas, traffic congestion is often self-limiting. Especially during peak times, for every one person on the road, there are many more at work or at home waiting for the congestion to clear up before they head out. This latent demand means that any increase in capacity will quickly be filled up with more traffic, bringing the congestion back to the same level it was before. However we accommodate it now or in future, traffic will continue to be one of the biggest challenges in our urban areas and traffic signals will continue to be one of its solutions. Thank you for visiting, and let me know what you think!
Instagram Engineering
Shopify Engineering