Elders, Wellbeing and Learning By Doing (2/2)
Part 2: The Grandma Phone
In the first part of this series I exposed you to the big challenges I have been facing (elderly, isolation, relationships and access to technology) along with the importance of interdisciplinarity and a good ecosystem when doing so.
This time, I want to immerse you in the details of the design and prototype phase of the device I built to potentially help tackling those challenges. Next, I will explain you how this project is also a “Learning By Doing” project, my takeaways and the next steps.
Design and Prototype
After all the framing, it was time to get my hands dirty. For the scope of the course, I decided to center my energies on designing a solution that would tackle elders needs first. As future work I would expand it to other groups: people with reduced mobility and people with disabilities. The decision was taken out of pragmatism as I could more easily think about my grandma environment and needs than anything else.
Before any brainstorming session I decided that the device would have a few but relevant constraints:
- Low cost: I want to build a device that is affordable to anybody willing to acquire it or set it up. Ideally below 50 euros.
- Open source: Share the codebase and allow other makers/developers to adapt the solution to their communities or relatives needs.
- User centered design: Put the user in the center of the design decision making. This is especially relevant targeting a sector of the population that is generally reluctant to adapt to new technology.
Ideation
During the ideation process important questions were raised: what would be the most natural way for an elder to communicate with a device? Where would it be placed? Would it be a wearable? A device sitting in the living room? How can we design taking into account previous habits and knowledge?
All these question were put in discussion with peers, mentors, family (included my grandma) and as many people as I could share it with. I also did a research on current products tackling the same problem such as this, this and this one and analyse how they could better serve their purpose. The sketch of first solid idea looked like this:
In words, the current TV, sitting in the living room, would serve as the interface and audio support for a video-calling app that would be streamed to the TV through a little and low-cost computer (Raspberry Pi) plugged-in to it. The user would navigate through the app interface using the TV controller and a video-camera would be connected to the computer so the user on the other end of the communication channel could see the person talking.
I liked this idea because it is re-using a device most elders have at their places and also it is taking advantage of previous knowledge: they know how to use the TV and a TV controller but…“Wait a minute! This a totally new and active way of using the TV controller!”. That’s true, but I assumed that the learning curve would be considerably shorter than learning how to use a new device from scratch. Moreover, any TV model, no matter how old or new, could be used to plug-in the computer that would be streaming the video-call app. Neat, huh?
I naively called this invention the Grandma Phone, which almost sounds like gramophone.
Iteration
I was pretty satisfied with the solution but after few conversations I questioned if the TV controller was really the best way to interact with the app. In the first place, the TV controller has not been designed with the purpose of controlling a very specific app and it could have limitations to proper navigate it with ease. So what else? What’s the most natural way humans have to communicate? Language! Could I integrate a voice-assistant to the solution? Yes, why not!? A voice-assistant would create a seamless and the most frictionless interaction with the device. Users would just need to learn the proper commands to communicate with it.
Even if it looked like a good improvement I wanted to make sure I was not over assuming things. Interestingly, a research was conducted to explore whether voice interfaces would be prefered by seniors compared to interactions using physical interfaces such as mobile phone keypads, touchpads and computer mice. The research showed that the elders in the study had a very much positive experience with voice user interfaces and highlighted the need to provide alternative channels of interaction if the voice user interface does not do what the user expects.
Scope and feasibility
Great. Time to create! After sorting how the device would be and how the interactions would be handled, it was the moment to make it a reality. There were 3 main parts to develop:
- User Interface
- Video streaming infrastructure
- Voice-assistant
After a preliminary research on how to implement the video streaming infrastructure, I dismissed to include it as part of the scope of the “Learning By Doing” course. The time to deliver was limited and what mattered the most was to show a prototype that when seen could give a clear understanding of its purpose and its interactions. Therefore, I decided to focus on the most crucial part: the interaction pipeline of the user. From speaking out simple commands and the needed guidance using a voice assistant, to the interface that would react to the commands and show the proper and expected information.
To the unknown
I have the tech expertise to implement the interface and create a communication pipe with a voice-assistant but I had no idea how to create a reliable voice-assistant and how to design an interface that would suit the needs of elders.
Thanks to the Internet and the distribution of knowledge there is an increasing amount of high tech software available to developers which they can easily integrate to their own projects at no cost. That is the case for voice-assistant tech and among the many voice-assistant solutions provided by different companies I decided to use SNIPS, a privacy first voice-assistant. One problem down!
However, I still lacked the required knowledge to design interfaces adapted to elders needs and it seemed a pretty niche domain to actually find somebody who could give me a hand. Therefore, I decided to find that knowledge in the Internet and learn whatever I needed.
After going through many articles and research (find my notes and the articles I read here) I extracted few guidelines that would help me to direct the design of the interface:
- High contrast between the foreground and background
- Avoid animations and moving text
- Strive for consistency
- Icons accompanied with text
- Give ample time to read information
I designed three views: contact list, calling contact and on going call.
I used Sketch and afterwards InVision to have a feel on how the interactions would occur. Also, that allowed me to show a materialised idea and get feedback before coding it. The last piece before starting the development of the device was to design the voice interactions the user would have when talking to the voice-assistant. Taking into account the views designed, I thought about 2 basic scenarios: call a person in the contact list and stop the call. The former scenario would have an extra interaction if the contact you want to call is not in the contact list, moment where the assistant would inform you about the situation and give you the opportunity to call somebody you actually have in the list.
The implementation took place in a span of a week, you can check out the code in Github. I can not find the words to express how exciting was to see all the pieces fitting and finally having in my hands what I have just thought about a couple of months ago.
Afterthoughts
By the end of the course I produced a device that can potentially help elders without computer literacy and people with reduced mobility or disabilities to unlock access to technology. I think this device could even be a way for elders to lose fear when using new technology, approaching interactions with voice makes the entry barrier very low. The later also means that elders have the opportunity to experience and learn about the intrinsic logic of modern but yet usable user interfaces which may help them when operating new user interfaces they have never seen before. Lastly and most importantly, I hope the project developed, in its further and more mature stages, can help break isolation and improve the quality of relationships and life for elders and people with reduced mobility or disabilities.
The importance of the ecosystem (again)
Although I am satisfied with the final result there has been a lot of relevant knowledge I have not gained access to and that would have make the solution more holistic. Perhaps, my approach on how to tackle the issue has been completely erroneous because of missing proper context. For instance, it would have been highly valuable to reach out caregivers and get their inputs about working with elders. This shows that even in a resourceful environment as in the CRI, there will be always limitations to access expert knowledge. I wonder how my prototype would have ended up looking like if I would have started designing it in another context with way less opportunities to valuable inputs and interactions.
When you are dealing with unsolved and big problematics you will need to view the problems from as many angles as possible to see what no one could see on her own.
Takeaways and Learning By Doing
After all, how is this project a “Learning By Doing” project? In it is more simplistic definition “Learning By Doing” means to learn knowledge by hands-on experience. It means active learning through the actions the learner performs. There is no well structure instruction of how and in what direction the learning is going to occur. It’s important to highlight that the mere act of doing it is not enough if what is done are the same things over and over again without getting external positive or negative feedback. This feedback encourages the learner to critically improve the thing or things being done.
Throughout the article, there has been multiple examples of aspects of this project that relate to a “Learning By Doing” project. In the first place, the problematic definition that in its core was laying on a very personal level but that hid a much larger problem: computer literacy, wellbeing and inequalities. I had the change to deeply explore the problematic and understand it as much as I could through a critical lense. There was also the community aspect, sharing my ideas and getting feedback to validate, improve or discard them. This part proofed to be the most relevant one as it helped immensely to narrow down the form and shape of the solution and to come with something genuinely different. Once the idea was clear, I was able to start defining the design and technical specifications and figure out the things I didn’t know how to do. That involved research and trial and error. I had to prioritize the most relevant features to proof the concept of the device and finally get to the nitty-gritty and plan how I would be executing the implementation.
Moreover, the intrinsic motivation of finalizing the project brought me beyond some aspects I did not think about at the start like documenting and sharing the whole process as I did in these two articles.
If you put all this in perspective there are 5 key things I learnt or improve during this project: identify problematic and justify it through research and critical thinking; seek innovation through sharing and listening; dive to the unknown and figure out stuff; project and time management; document and share. The most beautiful thing is that none of those learnings were something expected and they were being defined on the fly as there was a need and interest.
Next steps
There is still a tone of work to do to make the Grandma Phone a reality. One of the first steps would be to start testing the prototype with real users and get a feel of how they use it and what could be improved. It would be very useful to create an on-boarding one-pager infographic with instructions explaining the basic interactions you can do with the device and what to expect from it. Finally, the next natural step would be to create the video infrastructure and the login functionality (how to login without an email? maybe use biometrics?) to the app as well as the app the more computer literate users would be using from their phones: relatives from the elder, more savvy friends, caregivers, etc.
If all of this rolls out successfully I hope to be able to package it as a product but also, as it is an open source project, to give the power to other developers to set it up on their own.
Long live to the Grandma Phone, the hyper connected grandma!
Pau is a master student at the EdTech Master at the CRI and a maker and developer interested in design, social causes and the way humans learn and document experiences.