Perceptive Pathways
A Safer Way to Cross the Street
Motivation
As students at the University of Pittsburgh, we have often found ourselves in dangerous situations when crossing the street. We heard multiple stories from actual crosswalk users about their close calls at crosswalks, and after conducting some research we found that nearly 6,000 pedestrians died while crossing the street in 2017.
Drivers and pedestrians are not always aware of each other's presence at crosswalks, which can lead to dangerous situations. This led us to ask ourselves, how might we improve both the safety and awareness of both drivers and pedestrians at crosswalks?
Human-Centered Design Methods
We began the design process by attempting to understand who we were designing for in the first place. We surveyed over 150 users of Oakland crosswalks using a location-encoded QR code survey. This allowed us to understand the problems faced by crosswalk users, as well as the areas with the most problems.
Additionally, we reached out to several experts in the field of pedestrian safety. We initially talked to Paul Supowitz, Vice Chancellor for Community and Governmental Relations at Pitt. He helped us to further understand the problem of pedestrian safety at crosswalks and informed us about current efforts for improvement on Pitt's Campus. He gave us several additional contacts, including Officer Guy Johnson, a Pitt Police officer and pedestrian safety expert, and Katy Sawyer, a project manager at the City of Pittsburgh Traffic Division. They offered us their insights into what solution areas to focus on and the feasibility of different systems we were considering.
Design Goals and Pretotypes
We started the design process with several goals in mind. We wanted a system that would:
​
-
Sense the velocity of oncoming traffic
-
Grab the attention of drivers when pedestrians want to cross
-
Have a simple user interface and not be bothersome to use
​
From there, we designed two initial experience pretotypes. One was a gate, intended for use at intersections with stoplights, that would prevent pedestrians from crossing at the wrong time or when oncoming traffic was detected.
We additionally built a simulated velocity-sensing alert system for use at yield-to-pedestrian crosswalks that would enable drivers and pedestrians to communicate, even in low visibility situations.
After some preliminary testing, we found that pedestrians would rather be empowered than impeded when crossing the street; the gate received overwhelmingly negative feedback, so we decided to proceed in designing a solution for yield-to pedestrian crosswalks that would include a means of communication for drivers and pedestrians and velocity sensing capabilities.
Our System In Action!
How It Works
System Overview
Technology Overview
The brain of the system is an Arduino Mega, running a modified version of "C" code. Inputting to the Mega is a large, light up red button which is pressed to start the system reading the vehicle speeds. Also inputting is a Bushnell Radar Gun, which we were able to splice and solder into a ribbon in order to properly wire into the Arduino. The Arduino interprets velocity values from the radar gun's LCD Display.
​
Outputting from the Arduino is a pedestrian walk sign that uses LED matrices to light up an overlaying cut-out in the shape of the desired output. Also connected is a waterproof LED strip which is attached to the walk sign via a strong adhesive.
Results and Future Directions
Due to time constraints and government regulations, we tested our final prototype with users in the Benedum Hall Parking Lot.
We were able to verify that our system was able to detect the presence of moving vehicles, and we observed a vehicle detection range of greater than 200 feet.
​
After testing with 29 test users, the majority of them reported that they would use the Perceptive Pathway if it were in place and would feel safer if the system was in place. The chart below summarizes our findings after testing:
Ease of Use: Users reported being able to easily understand what they were supposed to be doing and seeing
​
High Use Rate: We tested the button alone without any additional functionality, and the majority of passersby pressed it anyway
​
Weatherproofing: The system is almost entirely waterproof, except for a couple of unfilled wiring holes on housings for the button and radar gun
​
Alerts Drivers: We were able to use the combination of the large sign and LEDs to notify drivers of pedestrians
​
Alerts Pedestrians: The pedestrian sign was able to effectively communicate to test users whether it was safe for them to cross or not by communicating with the radar gun
​
For future iterations, we would like to integrate a solar panel into the system to enable it to power itself independently. Additionally, we would need to add additional wireless capabilities in order to scale the system to both sides of the street, as well as a sound system to accommodate for the visually impaired.
Meet the Team
From Left to Right:
Warren Lester, Tyler Zinn, Frank Czura, Brett Saunders,
David Garrambone, Joe Landsittel, and Lucas Bechtold