Inspiration
Safety-based applications currently existing in the market require the user's input to signal any emergency. We wanted to build an app that can passively sense when the user is in danger, without any manual input.
What it does
The idea is to incorporate sensory data from accelerometer, galvanic skin response system and heart rate sensors to identify symptoms of the flight or fight response that stem from fear. We laid our focus on three prominent symptoms - changes in breathing patterns, elevated heart rate and change in skin conductance due to sweating. Based on anomalies detected in the sensory data, further actions to alert emergency contacts or seek help are taken.
How we built it
The application module (apk) was built with Android Studio. Sensory data was collected from 6-axis accelerometer embedded on the Arduino 101 microcontroller to sense changes in the breathing pattern. This sensed data was relayed to the apk through the low-power bluetooth module on the Arduino 101. The apk processes the received data to detect any outliers as an indication of the aforementioned symptoms. Based on the detection, a first-level filter is applied through a security question that the user must answer if safe. In case the user is unable to respond to this question, the user's emergency contacts are alerted indicating a threat. User's emergency contacts can indicate their ability to offer help. If they fail to do so, the application alerts 911.
Challenges we ran into
As first-time hackers and inexperienced coders, we experienced trouble learning the documentation of the Arduino and android-studio packages. There were issues transmitting data with the help of the Bluetooth module. The android application frequently crashed as well.
Accomplishments that we're proud of
Ideation, learning two packages from scratch within a very short span.
What we learned
Programming for the first time on Android-studio and Arduino platforms, fundamentals of Java.
What's next for Periculum
- Completing the android application in progress to accomplish our fundamental idea of alerting contacts.
- Extending sensory information to the GSR system and heart rate sensors using photoplethysmography or ECG biosensors.
- Optimizing energy considerations.
Log in or sign up for Devpost to join the conversation.