DriveClear

Homescreen
Collecting user data
Reaction Test Screen
Calculator Screen
Speech test

Inspiration

8 teens die everyday from driving while intoxicated. We sought one more step that could come in the way of the choice to drive after a night of drinking--DriveClear. We wanted to make an app that would inform the user of the changes that have come since a night of drinking on motor skills and various other functions that are vital to driving safely. By doing so, the user can get an idea of how his/her driving may be impaired and how dangerous it could to be to drive.

What it does

DriveClear allows the user to take several sobriety tests to test certain motor functions like reaction time, balance, and speaking while sober. After these tests are taken, a baseline is set up for the motor skills of the person while sober. The same user, when possibly intoxicated by any substance, can take the same tests at a later time. These tests and the data received is compared to the baseline tests to give percentage differences. Additionally, based on the number of drinks consumed, gender, and weight, the BAC (blood alcohol concentration) can be estimated for the user to also decide if it is safe to drive.

How we built it

We first researched and brainstormed several sobriety tests that could be conducted using a smartphone. We discovered that people who are intoxicated would have trouble walking balanced, reacting fast, and talking without slurring. We formulated different ways to test these problems caused by alcohol consumption including an accelerometer walking test, react button test, and a speech test. We used IBM Bluemix and Android Studio to make the app.

Challenges we ran into

Some challenges we ran into included delegating work in the project. At first, we did not know how to split up each project, but after some discussion, we were able to identify the strengths of individuals in our team to split up different tasks that could be accomplished amongst us. We had trouble with the consistency of the eye detection test as well.

Accomplishments that we're proud of

We're proud of the real world applications of the product and the potential the product has to save lives. Our product, DriveClear can be used by students and adults worldwide to make responsible decisions when tools such as breathalyzers are not available. The app reinforces the truth behind loss in motor function when drunk as well, thus proving to those in need of making a decision on driving while under the influence. By choosing not to drive, the person under the influence is not the only person made safer, but also others that could be on the road.

What we learned

For the majority of our group, this was our first time using Android Studio to code apps, meaning this was our first experience with XML and Java in terms of Android. In addition, we have never worked with the IBM API Bluemix before, so learning how to implement it into our program was a new experience.

What's next for DriveClear

We plan on expanding DriveClear with possible more tests. We got a preliminary version of eye tracking finished, but it was not consistent enough for full release. In the future, we would like to add eye tracking to the sobriety tests. After our research, we found that eye movement in someone that is intoxicated is generally sporadic and by using the facial recognition API, OpenCV, we could detect those movements for any tests. Additionally, since we did not have enough time to make the UI look better, we could focus on making it more user friendly. Furthermore, we could bring DriveClear to schools and colleges to encourage the use amongst them so that potential dangers with driving under the influence can be avoided.

Built With

Submitted to

HSHacks III
- Winner Best Android App

Created by

I worked on the initial idea production of the app. I also worked on the utilization process when creating the app. I helped to implement the reaction + walking test feature for the app.

Vikram Shirsat
I worked on the speech to text using IBM Watson and handled serialization as well as general app UI and displaying results. We worked closely with the IBM team to resolve many of our issues.

Varun Tandon
First I learned how to use Andriod-Studio, then I worked on developing our idea and then coding one of the four sobriety tests that we used. I also helped in implementing my part of the program into the main program. Lastly, I helped to code the BAC calculator feature.

Vincent Li
I came up with the original idea that the team later built on. I worked on detecting unbalanced walking by analyzing data from the accelerometer and stitched together the app from each individual contributor's features.

Ian Hsiao
I developed the reaction test and generated the main UI for the app. I also contributed to the integration of IBM Bluemix, in particular the speech recognition.

Michael Chen