A scientific approach based on traffic patterns and commuter mobility can optimize bike lane planning

When it comes to opinions about bike lanes, few of us are stuck in neutral. Love them or hate them though, new research says a dose of scientific rationality can help locate them in the best places. Congestion is minimized while more people ditch the car in favor of emissions-free, two-wheeled commuting.

Working with two other academics, smart city researcher Sheng Liu pulled data and talked to city planners in Vancouver and Chicago to develop a model that can help municipalities choose optimal locations as they expand their cycling lane networks in response to growing demand.

The study is published in the journal Management Science.

“Our model provides a systematic decision-making tool for municipalities to design new bike lanes using existing data,” said Prof. Liu, an assistant professor of operations management and statistics at the University of Toronto’s Rotman School of Management. “It helps policymakers better quantify and evaluate the potential benefits and risks of bike lane construction. In particular, it can predict whether and where traffic will get better or worse and if emissions will go down.”

Bike lanes have taken off across North America, leading to reduced traffic fatalities, lower-cost access to private transportation and improved physical activity for cyclists. But, as many commuters have already concluded, “ignoring traffic dynamics when designing bike lanes can needlessly worsen congestion,” the researchers write. And there might not even be much of an uptick in cycling ridership.

Part of the problem is that city planners tend to rely on simplified planning approaches that can’t account for all the factors that influence the impact of locating a bicycle lane on a particular roadway, or multiple ones in a roadway system.

The researchers’ model uses a city’s traffic and commuter mobility data to predict how cycling and traffic congestion will behave and change according to where bicycle lanes are located. The model estimates how driving travel time changes according to vehicle volume and road features, the attractiveness of cycling or driving on a roadway according to predicted travel times and the presence of bike lanes, and based on all of that, plus a host of other variables, which roads in a network will have the most cycling use and least congestion if bike lanes are located there.

Applied to the city of Chicago, one of the most traffic-congested U.S. cities and where expansion of its cycling network is a major policy priority, the model estimated that adding 40 km of additional bicycle lanes in specific locations would increase cycling ridership from 3.6% to 6.1% in the city’s downtown, while increasing driving time by no more than 9.4%.

“As bike lanes expand, some roads may observe more congestion, and some roads may actually see improved traffic,” said Prof. Liu. “At the network level, we find that the overall travel time for all commuters is shorter under the proposed bike lane expansion plan. This implies lower emissions as well.”

Acknowledging that bike lanes can inspire heated debates, Prof. Liu advised that “we should let data speak and follow a scientific approach” to evaluate their effectiveness. “Simply taking out bike lanes from the streets would not solve our congestion problem and could likely make it worse.”

The research was co-authored with Auyon Siddiq of the University of California, Los Angeles and Jingwei Zhang of Cornell University.