What it does

Recall is designed specifically for those who have early to moderate Alzheimer's Disease. At these stages, memory and learning are impaired, but most speech can still be conveyed. While in this state, most patients often suffer from memory loss which can severely impact their daily lives and functions. These people forget where they live, who their friends are, and even what they did yesterday and what they have to do today.

This is a very serious problem and to solve this, we have Recall.

Recall is a personal wearable assistant for Alzheimer's patients that serves as a central hub for their lives. When the person wakes up, Recall lets the person know of their activities for the day and who they need to meet. But Recall also features the amazing ability to answer the questions that the person has. For example, the person can ask, "What is my son's name?", "Where do I live", "What are my interests", and so on. The assistant can also answer general questions such as "What is the weather today?" and "What time is it?"

The wearable device stays on the forearm of the person and features a status LED, a button to start voice input, and a microphone.

In addition to the wearable, there is a mobile app that a care-taker or loved one can use to fill in important information. The care-taker can enter in people the patient knows, what events they have on a certain day, their hobbies, where they've traveled, etc. This app makes Recall highly customization to the patient's needs and allows for someone to easily and remotely care for their loved ones.

How we built it

We used Houndify to power the text analysis and combined that with a Parse database to allow for extended functionality. We designed and 3D printed the entire casing for the device and integrated all the components into a nice form factor.

Challenges we ran into

The Houndify API had many quirks that differed across the SDK languages, but eventually we were able to use Python to achieve our tasks. One of our first 3D prints also didn't have enough structural supports, so we had to redesign a bit

Accomplishments that we're proud of

We were able to not only use the Houndify API, but were also able to create a superset of it that featured far more functionality via the use of a database that persistently stores the user's info.

Logo base created by Sergey Patutin from the noun project

Share this project: