"Sounding" Traffic

Inspiration

One of the most recurring dangers for the visually-impaired is at inaccessible crosswalks (i.e. those that don't give vibration or sound notifications when traffic has stopped.) No matter whether the person is accompanied by a cane or a guide dog, traffic intersections continue to be a source of terror for the blind as they primarily rely on hearing to determine if it is safe to cross the street. However, research studies that have been recently conducted demonstrate that listening for oncoming traffic is an insufficiently reliable method for safely crossing the street.

Further, the crosswalks outside Mount Holyoke College (where Hack Holyoke is held), on Rt 116, have recently been subjects of intense discussion after the recent death of a community member while they were using one of these crosswalks. Talks have been underway with the town and state governments to address the unsafe traffic conditions there.

But this problem will continue to occur in many other crosswalks around the world, for both blind and sighted people -- because various environmental factors as well as careless drivers may make truly reliable crosswalk technology impossible to achieve.

Our Project

We wanted to address these issues and return the power of safe crosswalks -- no matter the location, weather, visibility or accessibility -- to the pedestrian. We were inspired by the blind community and decided to use sound as a tool to predict the safety of a crosswalk. We created a device to sound a buzzer with a changing pitch to reflect the changing distance of approaching traffic, in order to better determine when it is safe to cross the street. By being attuned to the pitch of the sound, one can also detect if cars are decelerating as they near the crosswalk. Because the source of this sound is with the user, a blind pedestrian no longer needs to rely on their hearing to detect distant traffic at a busy intersection.

Approach

We used an Arduino Uno, the HC-SR04 ultrasonic sensor, and a buzzer to create this device.

Challenges

We were a team of three, from three different disciplines, who met at the Hackathon and decided to collaborate on this project. Between the three of us, we had attended only two hackathons - making us complete newbies to the hacking scene. A few hours into hacking, we discovered that the only sensors that the hackathon could provide for detecting proximity had extremely limited ranges. We decided to proceed with an ultrasonic sensor that one of us had brought along to the venue. We agreed to demo the project with this short-range sensor with plans to create a reliable prototype after the hackathon. We also had to ensure that the buzzer sounded a pitch that was changing at a rate noticeable to the human ear. Other challenges were in effective teamwork when morale was running low, and additionally in realizing how inexperienced we really were when we realized at the deadline that we didn't know how to submit code through github or devpost!

Future Work

In the future, we want to incorporate the sensor components that work reliably in the range of distances that would truly allow the pedestrian to make a safe choice - an infrared or ultrasonic sensor that can detect what's 200 ft away. We'll also move from an Arduino Uno to the SquareWear board so that this can be a handheld or wearable device (such as on the wrist). Since the SquareWear has a built-in buzzer, we only need to attach our sensor and this can be done using conductive thread. For a visually-impaired user, this device would be the equivalent to "looking both ways before crossing the street" as the sensor will make sure no cars are approaching dangerously from either the left or the right side. Ideally, we'll one day have a device that can predict accurately from the acceleration or deceleration of approaching cars whether traffic will stop sufficiently in time for a pedestrian to finish crossing safely.

Built With

• arduino