Alexa Memory

Inspiration:

People living with dementia often lose access to core memories about their loved ones. Having to ask repeatedly for reminders about who their family members are and key stories they shared can be dehumanizing and painful for their family. As students with grandparents who suffer from memory loss, we wanted to create an accessible tool that lets elderly family members remain connected to their most treasured memories and access key details about their loved ones, all without worrying about becoming a burden to their caretakers or family. We aim to preserve cherished memories and human dignity through old age.

What it does:

To address these challenges, we built Memory. This cutting-edge tech with a human-centered approach allows seniors living with dementia to converse verbally with our Amazon Alexa add-on. Our app consists of multiple features as outlined below: Memory: Prompt our solution with simple questions like “Remind me about my son” or “Where does my daughter live.” Going beyond simple facts, seniors can ask for reminders such as “Tell me a story about my son” or even something as nostalgic as “I want to hear my childhood song.” These facts, stories, and pieces of music come from user input, which can be custom-added by family members. Neuro-intervention: Decrease a patient’s agitation and improve communication and caregiver relationships - for instance, with music prompts delaying cognitive decline and promoting brain plasticity in the elderly brain.

How we built it:

Our application consists of a front-end (conversation) built with Amazon Alexa Developer Tools and a back-end built with Python and Javascript with AWS as our Cloud Server. We built our conversational recall assistant by designing and coding our own custom Amazon Alexa add-on and DynamoDB database within the Alexa developer console and AWS, enabling speech-to-text with accessible transcription of patients’ prompts, and comprehensive queryable insights allowing near-human-like dialogues between patients and Memory.

Challenges we ran into:

Creating the database from which memories and stories can be pulled was a major challenge. Learning to build our back-end on the Alexa Developer Tool, AWS, and DynamoDB was more difficult than expected, with limited documentation and experience building on the platform. Yet, considering the additional benefits of this platform and device - being easily accessible for dementia patients with its always-on capability - we considered building on this platform essential to enhance accessibility, thus tying our previous knowledge of Python and Javascript to Alexa Developer Tool extensions.

Another main challenge for us was our desire to learn deeply about the medical industry with a specific focus on dementia even with our 36-hour time crunch, meaning that we spent hours researching pain points in the industry. Despite the time pressures this introduced, it ultimately allowed us to refine our idea into a viable product.

Accomplishments that we're proud of:

As novice coders and first-time hackers, we’re proud of successfully learning and navigating an entirely new interface, creating a product focused on accessibility, and making a significant advancement for dementia patients. We are also proud of tackling a problem that is meaningful to us both on a deeply human level and that has broad-reaching socioeconomic implications. According to a study by Mayo Clinic, dementia impacts over 55 million elders across the world, costing caregivers and society over 1.3 trillion USD per year. While more than 60-70% of elderly people are affected by dementia today, only 30% have their condition under control. In a time of aging populations and shortages of young workers, it will become increasingly critical to elders to be able to recall memories independently and without reliance on caretakers. Harnessing the increasingly widespread technology of home conversational assistants like Alexa is, we believe, an important first step towards tackling this problem.

What we learned:

Technical: Building this app, we learned about new programming tools (Amazon Alexa Developer Tools and AWS), collaborating as a team, and creatively solving problems. Industrial: Throughout this hackathon, we also had the opportunity to learn about a significant industry and to learn from other developers and mentors who provided valuable industrial and individual pain points.

What's next for Alexa Memory

Beyond these 36 hours, we hope to explore several paths: We hope to allow Alexa Memory to record memories when elderly users do remember them and decide to tell their stories orally, so that when the user loses access to these memories, Alexa Memory can play it back to them. This feature would benefit not only the elderly but also their younger family members. The recording function can help to preserve oral histories that are often lost, so that they can be passed down through generations.
We need to conduct deeper research on effective behavioral treatments of dementia and the specific needs of dementia patients. We’ll do this through more interviews with professors and researchers, physicians, and patients and their families. These conversations will help us build product features that truly solve our users’ pain points and help us better empathize with their experience. Lastly, we aim to finish building out our user input system, in order to allow family members to input memories and facts that the elderly user can then ask Memory to recall.