Jul 16 2020
As part of a year-long project with Carnegie Mellon University's Master of Software Engineering (MSE) program, pathVu is developing pathCollect. pathCollect is a mobile app that will use machine learning and artificial intelligence to detect sidewalk hazards from smartphone imagery.
Imagine riding down the sidewalk on your e-scooter, bicycle, or wheelchair while collecting important data used to improve our sidewalks. With the help of CMU students, we will deliver pathCollect -- an application that allows us to map sidewalks from smartphone images collected every second."
Eric Sinagra, CEO/Co-Founder of pathVu
Four CMU MSE students, developing pathCollect during this three-semester long project in 2020, are working with pathVu engineers to solve a problem every city faces, hazardous sidewalks.
Initially, data collection will be done by trained pathVu technicians and will soon be open to being crowdsourced by the public. Once connected to WiFi, the data is uploaded to pathVu servers and is analyzed to identify sidewalk conditions. The images are processed through an advanced machine learning model that identifies tripping hazards and broken sidewalks and characterizes those hazards as minor, moderate, or major issues.
pathVu's expertise in accessibility tracks back to their research beginnings at the University of Pittsburgh and Human Engineering Research Laboratories. This partnership led to an ASTM International roughness standard, route accessibility index, and its engineer-grade pathMet tool. pathMet is used to conduct sidewalk surveys, using the accessibility index to help cities manage infrastructure improvements.
"We look forward to launching pathCollect right here in Pittsburgh communities," added Sinagra.