Aakash was in his Molecular Cell Biology class for Neuroscience when the professor mentioned Alzheimer's disease and the struggles patients of the illness face on a daily basis - forgetting family members' names, not recalling what happened in the past or what will happen in the future, not knowing where items are located, withdrawal from society, frantic behavior when alone, amongst other things. Dee and Aakash were avid Alexa users in their dorms and, realized that Alexa could be utilized to help patients with Alzheimer's manage their condition without needing constant medical attention.

function define_alzheimers(){
     console.log("a chronic neurodegenerative disease");
     var symptomsList = [
            "decreased and poor judgement",
            "withdrawal from work or social activities",
            "changes in mood and personality",
            "trouble understanding visual images and spatial relationships",
            "difficulty remembering fundamental aspects of personal life"

What it does

Our system has two services: an Alexa Skill utility and an intuitive, easily-accessible web portal for loved ones to access. People who know the patient well may log into the account and populate the profile with information specific to that patient - family tree (relationships and names), calendrical information, knowledge regarding where objects are stored in the house. Then, after saving the profile, the patient is able to ask Alexa questions about their own life as naturally as they could ask other people that they are close to.

"Alexa open my memory. What is scheduled for Tuesday?"
"Alexa open my memory. Where is my bread?"
"Alexa open my memory. Who is my mom?"
"Alexa open my memory. Tell me about my family."

It's as simple as that - loved ones manage a profile and an Alzheimer's patient asks the questions.

How we built it

Alexa Skill - We used the calendar example from the Alexa QuickStart Github repository as our scaffold and built off of that. After configuring the example project to fit our needs (changing names, adding a custom calendar from the web, setting up a web server on AWS, etc), we began implementing other features such as the family tree. For the tree, we set up a DynamoDB AWS system to store information on each family member for easy access later on. We used a similar system for the object/location database.

Web Portal - The web portal is written entirely in native Javascript, HTML, and CSS with Bootstrap for quick scaffolding, lots of sample images from online, and a hook up to the DynamoDB database.

Challenges we ran into

Wi Fi - Alexa relies on the internet to function properly and as we were at a hackathon with spotty Wi Fi, Alexa often struggled to get responses (or connect at all).

Lambda - Hosting the project in the first place was an issue because Lambda was down for our team for the first half of the hackathon duration.

NPM Problems - downloading the iCal module resulted in many issues so we needed to revert to a previous version and adjust the example code (aka mega-troubleshooting).

Accomplishments that we're proud of

First Alexa Skill Built - This is the first time any of us have built an Alexa Skill and it was a great experience! We plan on working with Alexa again in the future.

Real Life Utility - 48.5 million people in the United States were diagnosed with Alzheimer's in 2015. This Skill and web portal is definitely a utility that could change lives in extraordinary ways if implemented on a large scale.

What's next for Memory

We plan on adding an actual login system so people can create real accounts for their loved ones and begin using this Skill in real life situations. We will publish the Skill to the Alexa Store so people can download it for their Alexa and we will take advice and feedback to improve Memory for the long term.

Share this project: