The Madeira Touch: an Interactive Display

Robin is voice assistant for people with dementia or other cognitive impairments. It won second place in the 2017 CHI Student Design Competition. 

TEAM | Catia Prandi, RJ Villafor, Nicolas Autzen, Johannes Schöning

TOOLS | Sketch, InVision

METHODS | directed storytelling, competitive analysis, speed dating, usability testing, survey validation.

Above: a prototype of Robin's accompanying app.

Above: a prototype of Robin's accompanying app.


“Despite the numerous advances that technology has produced over the years, it often appears that swaths of the populace are excluded or left behind.” –CHI Student Design Competition

The 2017 CHI student design competition, entitled “Leveling the playing field”, asked student designers to find a population that was not currently being served by technology or a behavior that they wished to support. Based on the research conducted by me and Meg Nediver in our Check-In project, our team decided to focus on designing to support people with dementia and their caregivers.



Robin is a conversational user interface that supports independent living. It consists of a mobile/web applications that works with voice assistant devices to provide task prompts and guidance to users who need assistance. And although we developed Robin with dementia in mind, it could be used by anyone who could benefit from routine assistance within their home.


key insights

Dementia progresses through known stages.  We identified the first four stages of dementia as the most promising for in-home intervention, and found that in these early stages of the disease, many people are proactive and wish to learn about interventions that could help them maintain independence.


Routines are critical for living independently. Daily routines are widely used by people living independently with dementia, and caregivers do a lot to support adeherance to routines. Small deviations from routines such as missed meals or medications often led to negative health events, including hospitalizations.

But the exact manifestations vary a lot.  While routines are hugely important for health outcomes and life satisfaction, the manifestations of these routines need to be customizable. As the disease progresses, the assertiveness of the intervention needs to increase, following what one doctor called the “reverse Piaget” of dementia.

Ease of adoption and customization are crucial. Finally, we found that the rate of abandonment of technological assistants was high among our target population, making ease of setup and customization paramount.


3 types of assistance


enabling technology



Directed Storytelling and interviews


We returned to the same caregivers and healthcare experts that participated in directed storytelling for the Check-In project an conducted follow-up interviews.


Competitive Analysis

We reviewed assistive technologies that are currently on the market and technologies that are currently under development. We also spoke to people about what assistants they currently use, like pill organizers and emergency response devices. 


We read forum posts by people who had recently been diagnosed with dementia and recorded instances of needing help with certain tasks.

Forum research


Drawing on our scenarios from Check-In, we identified three type of tasks for which we could provide assistance: ensuring critical tasks are completed, providing step-by-step guidance, and quality-of-life suggestions.



We then created storyboards for each of the three types of tasks and speed-dated them with caregivers and healthcare providers. From these sessions, we grew more confidant that newly-diagnosed dementia patients would be open to trying a new assistive technology.

storyboarding and speed-dating


We performed think-alouds with Robin's mobile app using both low-fi paper prototypes and hi-fi digital screens.

usability testing


Finally, we returned to the dementia forums, this time asking posters if they would be interested in a voice assistant that could provide task prompting, routine guidance, and quality-of-life suggestions. 65% of respondents said that they would be interested in trying a system like Robin.

survey validation


concept video



See the paper in the ACM archives here