Next Project  >

signifiers.io

GOOGLE
 
guidelines for designing connected devices
ROLE
Design Lead

CLIENT
Google

DURATION
8 months

Smart technology will soon become ubiquitous within the home. But as these devices and the sensors within them become more complex, the harder it is for users to understand their capabilities. Increasing privacy concerns create a high barrier of entry for connected devices.


We studied how users understand connected devices and developed a physical design language to intuitively communicate device functionalities. Our findings are compiled into signifiers.io, a set of dynamic design guidelines for connected device makers. Our vision is to ultimately give users confident control over their devices within the privacy of their home.

Guidelines
In signifiers.io, we focused on how to design intuitive feedback and control mechanisms of sensors in connected devices. The website is meant to be used by future connected device makers, including project managers to engineers, while validating the design of a connected device.


The guidelines found on signifiers.io will allow makers of connected devices to create devices that allow users to intuitively and accurately understand how they function.

Our guidelines are organized by different categories of sensors and feedback mechanisms. Within each category, research is divided into general do's and don'ts, followed by more specific guidelines.
Screenshot
Each guideline is appended with justifications from our research to both support our findings and to encourage critical discussions for further investigations.
Justifications
Initial Research
& Scoping
For five months, we investigated the potential impact of IoT devices on privacy within the home, conducting expert and user interviews.
We took a breadth-before-depth approach to the broad challenge statement in order to learn as much as possible about the space of IoT and privacy.
 
Privacy is an area that was not commonly well understood and could only be accurately studied when looked at peripherally - otherwise, biases can easily and likely significantly skew participant responses.


Thus, in order to properly simulate the environment of a connected home in the future and study how people would behave, we chose to modify and conduct a diary study, run participatory design sessions, and speed-date assumption prototypes.

Diary Study
Diary Study
Participatory Design
Participatory Design
Speed-dating
'Speed-dating'
 
After synthesizing the information into insights, we discovered that a core cause for privacy concerns was discrepant user mental models of smart devices, creating high risks for privacy concerns when they learned that the device functioned in ways they didn't expect.
There's a widening gap between the capabilities of devices and how well people understand them.
 
We asked ourselves, how can people comfortably/confidently use smart connected devices within their home without needing technical knowledge?
Prototypes
We picked out several different sensor modalities like visual, audio, touch, motion, data transmission. Isolating them from the overall function of the device, we used techniques such as interaction metaphors, anthropomorphism, light indicators, etc. to illustrate a sensor's function.
Can we imagine ways that can intuitively communicate what the individual sensor does?
We came up with 30 clay and paper prototypes in the span of a week.
Clay Prototypes
 
When we consider the range of devices out there, we can get a sense of just how complex addressing privacy concerns will be. Each of these differences present unique challenges to our privacy and there are different design considerations associated with each of these contexts.
It’s important to look at devices because they’re the interface between humans and sensors; they provide the context for which we interact with sensors.
Building Process
Our prototypes were made out of laser-cut chipboard, popsicle sticks, Arduino parts, and a lot of hot glue.
To try to cover the widest range of devices and contexts in which they’d be used, we chose to build prototypes of 8 different connected devices. The number and diversity of devices we explored speaks to our goal of extracting insights to make guidelines that could inform the design of all connected devices in the future.


For each of these 8 devices, we made several versions of each one that featured the different feedback and control mechanisms that we explored. Over the course of two weeks, we built 44 operational prototypes. Here are a few examples:

Clay Prototypes
Voice-Controlled Home Automation
Clay Prototypes
Smart Toilet
Clay Prototypes
Smart TV
 
Testing
After building all these prototypes, we tested them by having people interact with them scenarios that simulated real life use cases.
We had participants pretend to use Skype on our smart TVs, enter a house with our smart door locks, and even pretend to go to the bathroom with our smart toilet.
Testing Timelapse
 
After getting their initial interpretation, we probed specifically on how well each prototype addressed 4 key variables that we identified as critical for the understanding of feedback.


These were 1) difficulty in understanding feedback, 2) noticeability of feedback, 3) confidence in understanding the feedback, and 4) how clear it was when the sensors were and weren't recording data.

The findings from our testing sessions were synthesized according to the type of device/sensor and the context of the home.
 
Although our guidelines can stand on their own, and make an impact by themselves, we have only begun to scatch the surface of research and design in this domain. Further research should be conducted to consider secondary/tertiary users, glanceability of feedback, a universal device design language, and other questions outside of our limited scope.
 
What I found really cool & Future explorations


Team @ California
Presenting at Google
Ideation Process
Privacy and IoT was a topic that I had very little knowledge about when embarking on this project. Attempting to gain traction in any specific area felt like information overload at every step of the way. I learned about privacy engineering and behavioral economics - concepts that changed the way I thought about trends in technology. It's indeed scary to think about the implications of the exabytes of data that will be collected about us, but it's just as exciting to think about the value these devices would bring to civilization (cue overwhelming thoughts about the future of humanity).


My team and I invested a great deal of time into discussing our approach and the kind of impact we wanted to make. We became deeply entrenched in this topic, so much so that one of the biggest challenges I faced was how to clearly and simply communicate such a complex/nebulous subject to the average person (it also tended to put people to sleep). The frustration of seeing so many ill-informed reactions and discrepancies between behaviors and attitudes towards the topic of privacy actually served as inspiration for the idea of intuitive device design.


Not having a specific user group and creating a design solution for people at large for a client that has so much influence on technology was an incredible learning experience. The moonshot mentality we were told to adopt fostered an insatiable hunger for experimenting and knowledge, and I fully intend on applying this attitude to my future projects.