Cracking Congestion, Mode by Mode Around the Globe
How ITS Grad Students Are Speeding up City Traffic
On a good day the 1.4-mile trip from Bancroft and College avenues to College and Claremont avenues south of campus takes about five minutes by car. But most days are not good, at least during rush hours, and the ride along the two-way street can easily take 20 minutes.
Why? Because traffic halts behind trucks unloading their goods, cars waiting to make left turns, or drivers attempting to parallel park in shopping areas.
Traffic also clogs up when drivers stop to allow pedestrians to clear crosswalks before making turns, or when they wait for bicyclists approaching on their right to pass before it is safe to turn.
Buses also cause backups when more than one arrives at the same stop, and there’s no room for the second bus to pull over.
As cities go, Berkeley is far from large, and College Avenue is a small example of multi-modal transportation congestion—a situation where bikes, cars, pedestrians, trucks, and buses are all competing for space on the same narrow ribbon of concrete real estate.
In some ways College Avenue, with its mix of bikes, buses, cars, and pedestrians, is not representative of most U.S. or European cities, where cars have long been the dominant transportation mode. Instead, as drivers give up their cars for bikes or the bus, its traffic congestion begins to resemble the problems confronting developing countries where motorization is increasing rapidly, squeezing traditional modes—bikes, buses, and pedestrians—off the road.
At the Institute of Transportation Studies an increasing number of graduate students are focusing on Urban Transportation Operations, a field of research that refers to the control and management of all transportation modes in cities for both passengers and freight, according to ITS Director Samer Madanat. A number of projects at ITS’ Center for Future Urban Transportation, which receives funding from the Volvo Research and Education Foundations, focus on the interdependence of urban transportation policy and technology, and use the understanding of that concept to devise sustainable transportation strategies with the ultimate goal of reducing congestion and pollution.
“
This is a growing field for us because research on topics related to urban areas is becoming more important as more than half of all people in the world now live in cities,” Madanat explained. Bikes in China are quickly giving way to cars or electric bikes, despite the government’s hope to move bicyclists to buses. Motorized transit is growing rapidly in India, but cars must still compete with pedestrians, bikes, animals, and carts on roadways.
“Important questions include how much street space to allocate to different modes, such as cars versus buses, how to coordinate various public transport modes, such as feeder buses with bus rapid transit, and where and when to allow trucks in cities,” Madanat added.
As second-year grad student Vikash Gayah explained, “At some level all of us are using our understanding of the physics of traffic to organize traffic and control operations to meet our goals—which are the most efficient, fair, or equitable, way to move people and reduce interaction.”
In some cases, adding new lanes or roads is out of the question. “You have infrastructure that is already there, and it’s very, very expensive to tear down a building to add another road,” added Gayah. “So we’re trying to use more efficiently what’s already there. Maybe minimal changes, things that can be implemented very quickly, very cheaply, and to more efficiently use 
the space that is available.”
Map of Chengdu, China: ring roads and trunk roads.
The issue of fairness is another consideration.
“It’s also a question of organizing the modes so it’s more fair to all the modes,” explained Eric Gonzales. “So that people can continue to ride their bikes or continue to take the bus rather than leaving them and going to the car, making China more like the U.S. which is so car dependent. We want people to use sustainable modes, but make it easier and safer for them."
Added Yiguang Xuan, “It’s not about punishing the car driver or forcing peopled to use bikes and buses. It’s more about finding designs and systems to make everybody better off.”
New Models for Moving People
Nine graduate students recently sat down with NewsBITS to discuss the projects they are working on within the field of urban transportation operations.
Yiguang Xuan is studying how best to exploit the potential of a single intersection. Using data from Chengdu, China, he is trying to determine how to organize pedestrians, bicyclists, buses, and cars to move through an intersection as quickly and efficiently as possible. “We want to understand how to get the most out of a single intersection,” he said, adding that once he’s developed a model it will likely be field-tested in Chengdu. Ultimately, however, the model should be applicable to other cities, such as those in India, where there are multiple modes of travel.
Vikash Gayah is looking at a slightly larger road network—“maybe four to nine blocks,” or what the students refer to as “mesoscopic”—as opposed to Yiguang’s “microscopic” research. Gayah’s goal is to describe traffic in this so-called “building block” or small component of the city in a way that can eventually be scaled up to accurately explain multi-modal traffic at a city-wide level.
Gayah and Xuan are attempting to develop sorting methods to move multiple modes of traffic—buses, bicycles, cars and pedestrians—through intersections as quickly and efficiently as possible.
Karthik Sivakumaran is designing a model for bus feeder service to connect passengers to the trunk lines or primary corridors that ring a city or run through it. (See map of Chengdu above.) In the Bay Area, for example, the distance between BART stops is large, so there is a need for feeder services to connect passengers to the transit system. A series of concentric ring roads surround Chengdu, and Sivakumaran’s challenge is to build a model that will help passengers get to the planned Bus Rapid Transit system that will travel those rings. Using Chengdu data he will need to balance the operating costs of the transit system and the user costs, which are based on access and waiting times.
Weihua Gu’s work involves larger travel corridors or trunk roads, such as the ring roads around Chengdu. His research concerns differentiated service, such as a combination of express and local service, on trunk roads. “Under some conditions, the differentiated service will be operationally more efficient than a single service with only one bus route with fixed stops.”
Beyond Planning for Cars
Eric Gonzales and Celeste Chavis are looking to provide better transportation on another continent.
Using a macroscopic fundamental diagram, the two are working with a land-use planning group from Columbia University to better understand the traffic network in Nairobi, Kenya, and how to improve what is currently chaotic.
“We’re trying to figure out how many cars you can fit on the streets in Nairobi, and how many people that network can serve,” said Gonzales. By modeling the city and identifying these macroscopic relationships they can determine when traffic flow hits a tipping point and becomes congested.
Currently 50 percent of people walk along Nairobi’s streets, while 10 percent drive cars. The rest rely on “matatus,” informal jitneys, often small Nissan vans that carry up to 20 passengers. Chavis and Gonzales are focused on allocating the street space to the cars and matatus.
Currently the matatus are blamed for causing congestion as they stop erratically to pick up and drop off passengers. City leaders would like to ban the matatus, but Gonzales and Chavis point out that this informal transit system carries far more people to their destinations than do individual cars. And sidewalks are lacking for pedestrians who walk long distances on unsafe roadways.
Gonzales and Chavis are providing a “macroscopic” view of traffic in a city by examining traffic throughout an entire neighborhood.
“Each street in the city individually has very chaotic traffic behavior,” explained Gonzales. “You don’t get much from it because there is so much variability. But if you look at the traffic in a whole neighborhood, it starts to behave much more consistently.”
Eventually their research should provide the land-use team with information on where new roads should be added.
Ilgin Guler is examining traffic in Amman, Jordan as part of a larger project undertaken with ITS to transform and modernize that city’s transportation system. Guler is trying to determine if adding a dedicated bus line to major roads in the city will move more travelers to their destinations more quickly, perhaps even those driving cars.
“We want to get the buses moving faster without actually making traffic for cars worse,” she explained. “There are dedicated bus lanes in the U.S. as well, so there are existing studies and data to look at. Basically, we’re using Amman as a test field again, like most of the other grad students to see what more we can get out of this situation.”
Battling Bunching Buses
Josh Pilachowski is focusing on how to keep buses from bunching—the all-too-common tendency of two or three buses to arrive at the same stop at the same time instead of remaining properly spaced along the route.
“This is an old problem, but remarkably little work has been done,” said Pilachowski. “We know why buses bunch, but there’s been little success in attacking the root problem—which is when buses are no longer in equal headways or spacings, the system tends to move to failure, and two buses show up together.”
Transit signal priority, limited stops, a hierarchy of buses, even adding slack time to a schedule may all lessen the problem, but once the buses “move off equilibrium” they can’t regain it.
But new technologies such as GPS and wireless communications can help make buses intelligent decision-makers, explained Pilachowski.
“We want to minimize the amount of information that comes to bus drivers because they have enough to deal with. Instead, we want the buses to do most of the decision-making and know when the bus ahead, say, needs to speed up and the bus behind needs to slow down in order to stay in equilibrium. Ideally, the system should be robust enough to heal itself.”
A solution for bus bunching will not only help booming cities in China or India. It would go a long way toward making bus riders on College Avenue happier.
So as students at ITS find solutions for moving people more quickly and efficiently on other continents, the models they construct will also find their way back home where Californians continue to lose time and tempers to congestion.
--Christine Cosgrove