Managing personal health and care is difficult, especially for older adults with chronic diseases and conditions such as Coronary Artery Disease or Type 2 Diabetes. For those who may have recently experienced heart attack, the path to recovery is daunting even with the support of health care providers and cardiac rehab is a crucial part of this process. The problem is, in clinic cardiac rehab requires significant time, effort, and planning for adults undergoing treatment. Imagine that you have committed your afternoon to commute to a clinic to for rehab therapy, but when you get there your blood sugar is too low and your doctor tells you it's not safe for you to continue. Frustrating right?

We imagine a new way - using personal health technology and a conversational AI we are enabling safe and effective rehabilitation and health management in your home. With Alexa and smart health devices, you can collect your records to share your at-home journey with your health care provider, receive reminders for any required medications, and exercise safely. SHAW-CR has the potential to automate processes of care, enhance remote monitoring, and provide more intelligent care services for older adults.

What it does

The Alexa skill utilizes a clinically defined safety protocol to monitor an adults’ health metrics, medication intake, and glucose levels to guide people through safe cardiac rehabilitation exercise. Using Bluetooth enabled devices, continuous glucose levels, weight, blood pressure, heart rate, and SPO2 levels are collected through AWS applications and data storage. Alexa uses this information to offer people insights about their health, with the initial focus on those participating in cardiac rehabilitation therapy. A (patient/health care provider) web portal allows for patient onboarding and registration to set up scheduled prescriptions and exercise routines which Alexa will provide reminders. Additionally, Alexa users can log their medication intake, exercise, and report symptoms they may be experiencing to enable enhanced and continuous remote monitoring for health care providers.

How we built it

Alexa – The Alexa skill (code) was developed locally, using Alexa endpoints to route intents and utterances through the Alexa skills console and forward requests to a local development application. This allowed for quick iterations and code development without needing to rebuild the model in the console. AWS Codestar was used for the pipeline to build and deploy new versions of the Alexa skill. The code was created using Node.js, and utilized Alexa APL, Dialogs, Reminders API, and Alexa permissions to enable the personalized functionality.

AWS App & Backend – The backend data streams for the Alexa application was created using AWS Amplify, Lambda, API Gateway, S3, and DynamoDB. AWS Lambda and API Gateway was used to connect to 3rd party API’s for smart health device data collection for individual users.

Challenges we ran into

Challenges arose in a few areas while creating the Alexa skill. For older adults, possibly those with limited mobility or visual problems, we strived to provide a choice of voice or touch interaction to enhance their experience. To enable simultaneous multimodal engagement, it required lots of work to ensure people could interact how they’d like to. Because each screen supports visual or verbal interaction, multiple possible responses had to be handled which created complexity to fulfill longer dialogs. Additionally, to provide a more seamless experience with medications and reminders, we had to find clever ways to auto populate medication reminders from back-end data sources. It was important to avoid as much information entry with Alexa and use real-time smart devices to provide up to date information and reminders.

Accomplishments that we're proud of

We are very proud that we were able to create a multi-modal experience to help people interact how they choose. We found success in using other AWS tools to set up profiles and integrate with external services outside of the Alexa skill.

What we learned

Never having developed with the Alexa platform, we learned a ton about how to create conversational dialogs, integrating Alexa with other sources of information, and found areas that we’d like to focus on to provide more personal interactions for people.

What's next for Smart Home for Aging Well - Cardiac Rehab

We will continue to build on the foundation that we have with our prototype application through testing and usability studies. We hope to target other features including health alerts, more personalized exercise coaching, and hope to improve the interface design for easier and higher quality interactions.

Share this project: