After extensive testing at the Chemical Engineering Lab at North Carolina State University, with help from professors, we developed and perfected our sensor This high-quality, cost-effective sensor accurately determines the concentration of ethanol molecules in the closed airspaces of cars from the driver's breath, sending an analog value of the driver's breath alcohol content (BrAC) to the device's motherboard. 

We capture an image of the driver's face and port it to the model to detect features of intoxication. After our presentation at MIT, we shifted from convolutional neural networks to “quanvolutional” neural networks (QNNs). QNNs use quantum circuits and qubits to capture intricate patterns of facial images through enhanced feature maps. This optimization boosted our accuracy to 99.22%.

After getting a classification value and a sensor reading, we utilize this novel algorithm to incorporate data patterns seen in DUI reports by law enforcement agencies. We use data such as driver age, gender, current day and time, etc.