Inspiration

When we found it that: -South Carolina has a higher fatality/mile traveled than any other state by a long margin -Anderson County has significantly higher crashes along its main road compared to bigger and busier counties. -Fatal Crashes in South Carolina: 924 -Motor vehicle crash deaths in South Carolina: 988 -Nationwide at least 36 kids die in hot cars every year -According to the Centers for Disease Control, 50,000 people are hospitalized each year, and 430 die due to accidental carbon monoxide poisoning.

What it does

-Able to recognize a Drowsy Vs. Alert face using AWS Recognition. -Prevents the driver from drinking/sleeping and driving. -Able to open the window if a child or pet is left inside and is unable to escape or open the window -Able to read Temp (F) and Carbon Monoxide in Parts Per Million -Multilingual support with all voice prompts in both English and Spanish.

How we built it

We used AWS for facial recognition, posture analysis, and image/data storage. We use a Raspberry Pi to take the picture, display user interface, and output sound to a speaker. We used two Arduinos as an easy solution of multithreading. One Arduino reads sensor values, while the other Arduino commands a motor.

Amazon AWS Services used

-AWS Rekognition -AWS DynamoDB -AWS S3 Storage -AWS Polly -AWS Translate

Challenges we ran into

Building with cardboard, Tolerance issues with 3D printing, embedded systems with watchdog timers, calibrating machine learning models with a limited dataset.

Accomplishments that I'm proud of

Connecting all of the sensors, project boards, and AWS into one system.

What we learned

Secure your password at hackathons,

What's next for IoT Automotive

Expanding

Share this project:

Updates