Talki Talkie
A Voice Agent for
Elderly Healthy Eating
Interactive device with voice agent 2019
/ Design Challenge from Archipel
How can the staff be supported to create a
“Gezellig eating atmosphere” and motivate the elderly in Archipel to eat well with the same manpower?
*Gezellig means comfortable in Dutch
/ Overview
This is a semester-long, service design, UX research, and product-system design project, collaborated with the nursing home, Archipel in the Netherlands. We focused on enhancing the whole eating process in a caring home because we wanted to give back the right for the elderly to handle more of their own life.
Feature
Talki Talkie is a product, which empowers the elderly to increase their authority of healthy eating. A service design blueprint and voice interaction flow are designed for finding their needs in caring home ecosystems.
Tools
Semi-structured interview, CUI, Qualitative Analysis, Adobe suit, QUEST Questionnaire
Our Team
Veerle van Wijlen
Client | Archipel
Award
Story
We worked closely together with the whole stakeholders, including the elderly, chef, food manager, and caretakers. Also, the selected technology is chosen to explore the future technology.
My Roles
User Research, UX Design, Evaluation,
Modeling, Prototyping (Hologram), Film, Hardware Integration (Arduino, Processing), Technology Evaluation, Voice Agent
Lectures
Yuan Lu, Xipei Ren
SOLUTION
Talki Talkie is a system with a tangible digital product that empowers the elderly in care homes towards choosing what they eat. Several features are included as shown in the video.
/ Design process
There are 2 phases and 8 iterations throughout the semester long.
GDPR - Ethics, Privacy, and Data
Every step of the design process is concerned from a personal level of willingness, how they would like to present or not present their data? We followed the European GDPR regulation and were reviewed by the Ethical Review Board of the TU/e.
/ Background
Studio Silver
Studio Silver squad in TU/e collaborated wit startups, institutions, and corporates to provide design cases. We believe “Instead of adding years to life, we add life to years.”
Literature & Anthropological Visiting
Aging Population with Lower QoL
The aging population is an occurring phenomenon, in the Netherlands is expected to be around 4,6 million elderly in 2040 (PBL, 2013). This rise concerns the caregiving industry and the elderly’s daily quality of life (QoL).
About Archipel
Students research and develop strategies, products, or services. The framing of the case is that the staff needs to support a group of elderly with various diseases (dementia, brain damage, and etc.) to have meals on a daily basis.
Empowering the elderly
Various elderly’s eating problems make elderly’s greater dependency on caring staff. In order to provide quality care, we should make sure older adults keep in control of their lives (Tu, Wang, & Yeh, 2006).
Phase 1 | Iteration 1-2
RESEARCH INSIGHTS
Through literature research, contextual inquiry, a hypothetical user journey map,
and a co-reflection workshop, we figured out the importance of
Promoting fun social interaction between caretakers and elderly healthy eating process through a simple and efficient tangible-digital product service system
The following paragraph will introduce gradually.
Contextual Inquiry
As shown below, we did observations in nursing home environments,
and meet with experts (up / caretakers in the nursing home, food technologist)
After visiting the long-stay department, we found that empowerment to gain authority in choosing meals is the major task. Thus, the design challenge was
re-defined and tailored.
Brainstorm Mindmap
We did brainstorm again with the topics we concluded from literature,
interviews, and contextual discoveries.
/ TouchPoints from User Journey Map
Phase 1 | Iteration 1-4
After contextual inquiry, several interviews, and workshops with the stakeholders, the user journey maps have been iterated 4 times.
Stakeholder Map
Stakeholder Maps drew out the important stakeholders to research are caretakers, catering company and the elderly. It aims at visualizing the specific service, which provides the elderly healthy food, including caretakers, chefs, the elderly, and managers.
USER JOURNEY MAP
Besides understanding from the stakeholders’ perspective, we also drew out the hypothetical user journey map from the elderly’s view while we were observing.
Opportunity
A digital platform might provide the elderly a place to express their need in public. Also, through it, they have a higher chance to have personalized interactions with related stakeholders, such as caretakers, chefs, or food managers. Afterward, we validated the map with related stakeholders with a co-reflection workshop.
Mid-Term Proposal
Different concepts were proposed, including food shopping through VR technology, recipe co-creation game, and others. The proposals below were discarded as the statement.
Pain
Points
Lack of social interaction
and timing to increase the fun time
↓
Concetps
Gardening Table
Elderly has an “I don’t care attitude” towards choosing their meals due to lacking of time of caretakers
↓
Meal Co-Creation
Experiment with the potential benefits of technology as it holds the benefits of intuition
↓
AR Technology with hologram for interaction
Discarded
/Continued
After Interviews
The concept was not chosen since the elderly has to concentrate while eating more than normal adult.
The concept was supposrted, but the interaction can be further discussed and iterated.
Hologram has challenges with the light, weather, and increasing the workloading with the caretakers.
In conclusion, we decided to keep the design concepts of choosing meals together,
but changed the technology.
Co-reflection Workshop
Furthermore, we hold co-reflection workshop, which was aiming to look for various perspectives with stakeholders, like caretakers, food company and a chef in Archipel.
They provided the values of elderly healthy eating,
which are a good taste of the food and cozy or nice decorated environment.
Besides, we validate ‘food taste & smell’ with the elderly.
However, we discovered that the smell system of the elderly has degraded. Therefore, most of the elderly were unable to smell the garlic during our experiment. (The smell was strong enough that the caretakers at the next door came to ask what happened.) In this iteration, we decided to discard using smell as part of elements in our design.
/ Design Value
Phase 1 | Iteration 1-5
We concluded a new design value after the iterations in the 1st phase.
Let the elderly
see the transparency of the food production,
their voice gets heard,
and have the feedback in time from other stakeholders.
/ Final Design Concept
Phase 2 | Iteration 5
In the second phase of the design process, there have been three iterations, iteration 5 until 7. This phase was mainly focused on design, including interaction, dynamics, technology prototyping, and the service blueprint.
To validate the final concept set
To conceptualize the final physical design (Talki Talkie) and the service blueprint
Demo prototype and technology realization
Final validation with elderly residents
PERSONA
Through the previous iterations, we identified that the elderly lived in the long-term department is our main user.
Main Goal
A new round of ideation was done to solve 3 main problems amongst elderly in nursing homes regarding choosing their meals
1) “I don’t care” attitude
2) Elderly’s meal preferences being ignored
3) Elderly’s complaints after eating meals.
Ideation
15 sketches were gathered and evaluated based on their interaction complexity, technology feasibility, and level of intuitiveness.
4 final concepts were validated among the elderly for interaction style and roles of different stakeholders in the design. These concepts were designed based on choosing together.
/ Concepts Video Validations
Phase 2 | Iteration 5
In order to validate these 4 concepts with residents at Archipel, low-fidelity paper prototypes were made and filmed with 4 videos.
These videos allowed us designers to explain the workings of the concepts and ask specific questions related to interaction, technology and context.
FINDINGS
The most important insights from the elderly are as below.
They didn’t want to choose together with other people, as they have to confront with others, which influences their decision as now.
Interaction with physical design was preferred over digital interaction, because digital could be impersonal and distant.
Speech was more nature interaction than tokens, also they found token childish.
They want ‘Voice’ to convey their choice and opinion in the design.
/ Final Concept & Service Blueprint
Phase 2 | Iteration 6
After validation, new concepts were made implementing voice recognition, sound playback (talking function), physicality and personalized context.
Service Blueprint
We defined the service blueprint for visualizing the flow of interaction between Talki Talkie and the elderly in the nursing home space. Therefore, it clearly points out all its main functions, and the locations it might happen.
Food Company Visiting
Due to a lack of trust in the quality of the meals, we found there is a communicational gap between food company (Culivers) & caretakers, food company & residents. Through the visit, we extended from the ‘food company actions’ to embed into notification for increasing the trust.
CONVERSATIONAL DIALOG
Since the focus of the design was on the choosing process, a choosing flow chart was designed to prototype the dialogue with the elderly’s residents involving them and offering them opportunities for opinion recording.
Interaction
/ Demo Prototype
Phase 2 | Iteration 7
Light up
Notify of different phases of the process of food production and transportation.
Head (red) light up
Talki talkie starts to record the opinions of the elderly and the light lights up to inform the users.
Pick up
Play notification of food process and ring light stops breathing.
Platform
Other stakeholders could reply through the platform by texting, which will be played through talki talkie.
Hardware Design
All interactive and experiential functions were programmed using a variety of programming languages as Processing and Arduino; and communication tools such as web sockets, serial communication and libraries.
/ Validation
Phase 2 | Iteration 8
QUESTIONNAIRE - Technology Acceptance Questionnaire & Qualitative Questions
This final design of Talki Talkie was validated with 3 elderly residents (A / B / C), in their personal rooms (fig.14), and 1 caretaker (D), in the living room.
+
‘Talki Talkie is useful, faster, and easier for communicating with the caretakers and food company’ (A/B/D)
Talki Talkie is not really needed’(B / C).
The validation also made clear that participants were not sure how the design should work in long-term, but were open to try it out in the future, giving opportunities for future work.
/ Reflection
Phase 2 | Iteration 8
When choosing the technology implementation, we decided to choose a voice agent instead of a portable and elderly-friendly online platform for exploring new technology.
It’s a pilot design which use VAs as a mian technical assistant. It has a high potential to solve the daily interaction and reduce the workload.
Different scenarios are worth to be validated, such as elderly being along at home.
The validation also made clear that participants were not sure how the design should work in long term, but were open to trying it out in the future, giving opportunities for future work.