Efficient productivity is something we aim for, but it's often hard to figure out when we are biologically most ready to be productive. Implications from this lack in congruency include hours of inefficient work, restless sleep, and even conducting life-threatening activities under low alertness levels. We wanted to find a way to bio-metrically measure how alert you are, which will give insight to when you should be conducting different activities. Our device will not only improve productivity and efficiency in your life, but hopefully even prevent you from life-threatening situations by making you aware of your biological alertness.

What it does

We are using three different sensors (IR sensor, Galvanic Skin Response, and Pulse Oximeter), and using them to proxy your alertness level. Various scientific papers have implicated higher frequency of blinks, lower skin conductivity and lower heart-rate to lower levels of alterness. We algorithmically combine these data to give your current "human battery percentage," which is transferred onto a web app via a wifi-module in real time.

How we built it

We bought the three sensors ahead of time. We interfaced the sensors with an arduino to identify blink rate, skin conductivity, and heart rate. We are also able to analyze the data in MATLAB for more accurate real-time processing and visualization. We then feed all three data from the blink rate, skin conductivity, and heart rate in to a logic level shifter so that the proper voltage signal can be sent to the ESP8266 wifi module. By utilizing this wifi module, we are able to send the three biological data to Google Firebase Server live. This connection is maintained consistently to check for any other further updates coming from the wifi module itself. Lastly, we created a website using the latest javascript framework called react. This website keeps track of any changes in Google Firebase Server to display not only the blink rate, skin conductivity, and heart rate but also recommendation on whether one should sleep, study, or workout.

Challenges we ran into

Many. Getting sensor data and identifying what it meant. Calibrating sensor data. Getting consistent data with high signal to noise ratio. Wifi module: trying to connect to internet in general; weak wifie; biggest problem: sending the Wifi from wifi module to google's firebase; getting data from three sensors and trying to convert from 5V to 3.3 volt into the wifi module.

Accomplishments that we're proud of

Despite the challenges, we were able to get a working prototype. We were able to get around most challenges.

What we learned

We had not worked with interfacing matlab with arduino before, or used a wifi-module + all its implications before so it was a good learning challenge.

What's next for LifeBattery

We will work on getting better signals; having more training data sets to create more accurate results; we will get more sensor data; we will work on integrating with matlab better to give more instantanous data.

Built With

  • ardunio
  • electrodermal-activity-sensor
  • esp8266
  • firebase
  • logic-level-shifter
  • matlab
  • max30105
  • pulse-sensor
  • react
Share this project: