top of page
title background3.png

Memory:                           
Machine learning for the elderly

In the reflection of apply hi-tech to real life, I looked into minor groups of people - the elderly, and exploration on how they can benefit from machine learning in psychological aspects.

Design Brief

Responsibilities and Skills

Product Designer
User Research
Quick Draft 
Mock-up and Testing
CMF
Digital 3D Development

This is project focuses on care for elder people. Through research, I find that although there are diverse products in the current market which do aware of the inconvenience of elder people, however, there is little attention paid to their psychological aspects. Therefore, from the perspective of the users, I designed a interactive device which based on machine learning and can provide more emotion care for the elder.        

Project Type

Timeline and Teams

2018.
Solo work

thesis project

General Background

Asset 4_4x.png

In cities,  over 50% elders live alone  or with their  couples.
We call them  “empty nest”.*

Asset 5_4x.png

Over 20% elder people aged more than 60 are suffering from mental or neurological  disorder.*

Asset 6_4x.png

Mental health has an impact on physical health and vice versa.*

Asset 7_4x.png

However, mental health problems are always ignored by doctors and family. Elder people usually do not tend to admit it. 

Observation

I spent two days staying with my friends' grandma who live alone. Here are some moment inspire me most.

Asset 9_4x.png

Most time of the day, she just in a chair with silence or listen to radio.

Asset 10_4x.png

Except for chating with her sister who is also live alone, she has no chance to communicate with people.

Asset 12_4x.png

she collects old photos carefully and always take them out to have a look.

Asset 11_4x.png

As we get familiar, she shares a lot of memories with me, from her primary school to now.

Survey about those behavious

To confirm those behaviours, I prepare four problems and chat with 40 elderly people met in 2 parks

Asset 13_4x.png

How do you think of your life? full of variety? simple ? or is it boring?

Asset 14_4x.png

Do you often have chance to talk to others?

How often do you watch old photo?

Asset 15_4x.png
Asset 16_4x.png

Why do you watch old photos

Journey map: How an elderly living alone passes her daytime

交互17all wix-17.png

What people say VS What Data says

Asset 18_4x.png

It’s unecessary for elderly people to share things about past with their families.

Asset 17_4x.png

The following is a psychological experient which focus on depression elder. The comparison group utilized the traditional treatment methods. On the contrary, the experimental group made use of old objects to help awake the awareness of the elder about the past. here is the result. Y refers to the degree of depression level. The higher the number, the more negative feelings they hold.

But look into past with the company of old photo is so good for elderly people’s mental health.

How might we

Create a warm heart experience for elderly people who like to watch old photos?

What they need

Asset 19_4x.png

Sympathetic

Seeking resonance is human instinct. Old people living alone have few friends, there is little information that resonates with them. They are eager to resonate.

Asset 20_4x.png

Companion

The reason why elder people often turn on thee radio and TV when they are at home lies in that they hope to get a sense of companionship. 

Asset 21_4x.png

Fresh information

Monotonous life, boring TV shows make them more lonely. Fresh information can also help their mental health

How we do

The device can automatically identify the life tracks of the elders and then help to find friends who share the same experience with themselves. 

Asset 22_4x.png

Elderly people put photos in a “box”

Asset 24_4x.png

This box will play the photos as well as the  corresponding  songs of that age

Technology support: Machine learning

Public sources can provide large numbers of digitally archived photos taken over the past century with known dates, it’s relatively straight forward to obtain labeled data (dated photos, in this case) with which to train this network. Once the network is trained, we can feed in a photo, and we get out the year in which the system guesses it was taken. 

Asset 25_4x.png

Input image

Convolution

Max pooling

“Fully Connected” Network

Output

2D Development

交互20all wix-20.png

Mockup testing and feedback

In order to make sure whether my design was in line with the ergonomics principles, as well as whether the buttons are comprehensible enough for the elder to use, I made a paper model. During the test, I told the grandma that it is a device which can automatically play music through identifying the information behind photos. Then, I observed the methods she used to put the music on and other actions.  

Asset 28_4x.png
Asset 27_4x.png
Asset 26_4x.png
Asset 29_4x.png
Asset 32_4x.png

she quickly found the mouth and put the photo inside.

Asset 30_4x.png

However, she did not know how to make the music stop

Asset 31_4x.png

She thought the mouth was little bit wide. 

Asset 34_4x.png

The switches should differ from each other

Asset 36_4x.png

The surface should be more slope

Asset 33_4x.png

The mouth is too loose

Asset 35_4x.png

The buttons for volume and channel turning can not differ from each other

Ideation

Asset 38_4x.png
Asset 37_4x.png

CMF inspiration

3D development

Asset 1_4x.png
Asset 4_4x.png
Asset 5_4x.png

Volume

Asset 6_4x.png

Channel turning 

Asset 7_4x.png

Switch

Asset 8_4x.png

Previous

Pause / play

Next

When the photo is put, the music will automatically play and the mood Lighting will be turned on

Asset 3_4x.png

Storyboard

Asset 42_4x.png

1. An old man living with himself is looking outside the window with a sense of loneliness. 

Asset 43_4x.png
Asset 44_4x.png

2. With the missing to the past, he puts the photo into the device.

3. The device automatically reads the information and then play the musics of the corresponding years.

Asset 45_4x.png

4. The devices analyzes the living track of the old man and finds there are some elderly people sharing the same experience with himself. The community will organize an activity to enable them get to know each other.

4. The devices analyzes the living track of the old man and finds there are some elderly people sharing the same experience with himself. The community will organize an activity to enable them get to know each other.

Asset 46_4x.png

5. The old man decides to go out of home, and participate in the social activity.

Asset 47_4x.png

6. He makes some new friends and feels no more lonely. 

bottom of page