Data analysis

Pittsburgh Targets Bicycle and Pedestrian Infrastructure Spending Using Traffic and Accident Data Analytics

This audio is generated automatically. Please let us know if you have any comments.

A new approach to analyzing traffic data has given Pittsburgh valuable insights into how it can make its streets safer for cyclists and pedestrians. The city has contracted with StreetLight Data to measure automobile, pedestrian and bicycle traffic. StreetLight then overlaid the data with maps showing where bicycle and pedestrian crashes occurred to give the city a better idea of ​​the priority of traffic safety measures among the city’s 1,300 miles of major roads.

“Remarkably, busy shopping, retail and educational hallways were not necessarily correlated with crash severity. In fact, the most serious crashes occurred in areas with low bicycle and pedestrian traffic,” says StreetLight’s report on the project. Half of fatal pedestrian crashes and 51% of total crashes during the study period occurred in neighborhoods with low pedestrian activity, it says. For cycling crashes, it found a lack of adequate cycling infrastructure, as well as low cycling activity correlated with half of fatal crashes and 36% of total crashes.

The project used data collected by StreetLight, mostly in the pre-pandemic years of 2015 to 2019, which shows where people were riding and walking. StreetLight tracks movement via data from cell phones and GPS navigational aids to determine when and where trips start and end. It can largely indicate the type of movement based on speed, the company said. Even if the car traffic is moving at a rush hour, it can still estimate whether it is a bicycle or a car based on the travel time. This method doesn’t count people who don’t use cellphones or other trackable devices, but it can provide months or years of data from multiple sites. StreetLight said that when it compares its data with data collected by other methods, it correlates quite strongly.

Pittsburgh previously relied on old-fashioned data collection methods, such as using street rubber hoses, cameras or people counts to determine traffic volume. Its resources only allowed the city to collect data on a few sites at a time and for only a few days. “It’s difficult to collect field data for 10 to 20 intersections at a time,” said Panini Chowdhury, senior transportation planner for the Pittsburgh Department of Mobility and Infrastructure. “It was very time and labor intensive.”

To plan for safety improvements, “most cities look at where they see the most crashes. But you don’t understand [the picture] unless you look at traffic volume,” explained Martin Morzynski, senior vice president of marketing for StreetLight. “Some of the most dangerous streets are not the busiest at all”, they could have a dangerous level crossing due to an overpass or a blind curve. “Sometimes it’s those places that need fixing, like adding signals, a sidewalk or something to add visibility,” he said.

This data can help municipalities understand how to request and use the billions of new federal dollars coming years for transportation safety and cycling and pedestrian infrastructure under the Infrastructure Investment and Job Creation Act.

The data does not take into account factors such as limited road space, hills and road curves that Pittsburgh must consider when implementing safety measures such as adding bike lanes, signs stop signs and pavement markings. “Generally for Pittsburgh, the intersections are the most confusing part,” Chowdhury notes. Data from StreetLight provides the number of cars turning at a given intersection, which helps the city determine if it needs a right-turn lane, he said.

Since a city cannot solve all of its traffic problems at once, the data helps it “prioritize our project corridors where we see a strong [pedestrian]/cycling activity and therefore a good place to full street development,” Chowdhury said. When combined with what the city already knows, the data from this research helps it choose residential streets that will get traffic calming devices to slow traffic and give priority to pedestrians, cyclists and to other non-vehicular vehicles through its “Neighborhoods” program. The data can also tell the city if many motorists are passing through these residential areas. And when the city learned where most cyclists were going, it rolled out bike parks in those areas as well, StreetLight said.

“We have nearly two dozen projects in the design phase that will be completed this year,” Chowdhury said.