Select Page

UMass researchers are using drones to learn about traffic risks for cyclists in Somerville

UMass researchers are using drones to learn about traffic risks for cyclists in Somerville

Researchers from the University of Massachusetts Amherst are using drones to examine the crash risks between cars and bicycles in Somerville. Last week, three UMass…

Tuesday, Oct 01

News

Researchers from the University of Massachusetts Amherst are using drones to examine the crash risks between cars and bicycles in Somerville.

Last week, three UMass students and a university drone instructor launched four drones from Conway Park. Their mission: collect video footage of a roughly one-mile stretch of Beacon Street. The project began amid calls for better bike safety following a fatal crash in Cambridge earlier that week.

Although the research was planned prior to the crash, Eleni Christofa, a professor of civil and environmental engineering at UMass, said the location was chosen because Somerville has invested in bike lanes and signage aimed at improving safety and comfort for cyclists.

“There’s a lot of bicycle traffic here compared to other places,” said Christofa, who is leading the research effort.

The drones will capture video over the course of several days, collecting data to help the research team understand how drivers and bicyclists behave around different types of bicycle infrastructure, Christofa said.

The footage will be analyzed by a Swiss startup company, MobiLysis , which uses drones and a type of artificial intelligence called machine learning, which seeks patterns in data. The company uses the technology to identify and recommend solutions for transportation-related problems, such as how to reduce pollution and ease congestion.

“We can assess the probability of collisions and see how safe the intersection is and if there is any measure that can be taken to improve that,” said Dimitris Tsitsokas, a senior data scientist and transportation engineer for the company.He said MobiLysis has […]

Share This