IBM Watson IoT and Bluemix Challenge @ VolHacks 2016

HeartTrends, an IBM Watson IoT Web Service that monitors patient heart rate and alerts families of abnormal readings.

About Me

My name is Christopher Pineda and I am a Computer Science undergraduate student at University of Maryland University College. I love building cool projects and solving interesting problems. When I’m not coding, you can find me doing school work or playing with my kids.

My Team

My friend and coding partner was David Campbell. We drove 8+ hours from the Mississippi Gulf Coast but it was worth it to attend our first Hackathon! We were meant to be teamed up with Aaron Eden, but he had to drop out at the last minute due to illness four hours before we left 🤒.

Our Platform

When we first arrived we weren't sure what problem to tackle or what platform to build off of. We made our rounds to met the event's sponsors and that is where we met Gayathri Magie, who is with the IBM Watson Internet of Things Platform. After talking to Gaya and completing a Watson IoT and Bluemix tutorial, we decided that it would be our platform of choice for VolHacks.

The Problem

The world faces an aging population. Advances in medicine, sanitation, and quality of life have resulted in increased lifespans throughout the world. According to the American Medical Student Association, the population of individuals over the age of 65 will increase by 73 percent between 2010 and 2030, meaning one in five Americans will be a senior citizen.

Our Solution

By providing alerts to emergency contacts when a patient's heart rate enters critical ranges, we believe that an IBM Watson IoT web service can save lives. We were inspired by The MIT Smart Mirror and the IBM Watson IoT Platform tech talk.

Implementing HeartTrends

We used the IBM Watson IoT Platform on Bluemix to create HeartTrends. We simulated data from a heart rate monitor and fed that into Bluemix. Once that data is parsed, we evaluate it against custom value ranges to determine if the readings are within a safe range. If the reported heart rate is above or below the designated safe range, we mark the reading as critical and begin the alert portion of our product. We trigger an HTTP request that pulls a JSON object from a database that contains the name of the patient’s emergency contact. We use this information to send the contact a critical alert using the Twilio API. Additionally, we record and visualize all heart rate data so that it can be reviewed by the patient’s healthcare team.

What's next for HeartTrends?

By adding authentication and a backend database, this product can be made to work worldwide. Additionally, there are several other sensors and devices that could be connected as inputs to IBM's Bluemix. IoT devices that implement GPS or RFID can be used to track the patient's location within their dwelling. IBM's Bluemix could watch this input flow for abnormal travel behavior or even the lack of motion.

Here are some other IoT inputs that we would want to tap into in the future:

  • Wearables to track weight and body fat percentage through electric impedance.
  • WiFi enabled thermostats, checking temperature to ensure HVAC is properly functioning.
  • Electrical outlet monitors that would identify patterns in electronic usage (e.g. TV, radio, cooking appliance, laundry).
  • Water guages that would identify the water usage of the patient's toilet, sink, and shower.
  • Pressure sensors for the patient's chairs or bed.

By aggregating the data created with these IoT devices, we can leverage the power of IBM Watson and other Bluemix services to create and react to a more comprehensive model of our patient's life. We would then also be predicting future needs that our patient may not realize that they have.

We’ll also be educating our employer about how we can use IBM Bluemix to radidly develop complex scaleable web services.