Redesigning a Thermostat: Final Thoughts

In Rapid Ideation and Creative Problem Solving, we invested eight weeks of time into redesigning a Honeywell Thermostat’s Interface.  On a micro level this has been about tweaking the placement of a lot of pixels from week to week and spending entirely too much time thinking about temperature.  However, on a broader scale what we’ve been learning through practice is how to rapidly articulate ideas, produce meaningful representations of them, and coerce meaningful feedback from users.

And all of those skills mean that we can confront a problem space and apply our unique frame as designers in a creative, iterative process that leads to more meaningful interactions.  So although I don’t care much about thermostats, I thoroughly enjoyed the challenges we’ve overcome in the last quarter.  And I ended up with a design that I’m proud of:

You can also check out the Wireframes via User Flows

Although I’m happy about where the design ended up, I tend to struggle with premature judgement while I’m trying to get to that place.  So I’d like to talk a little more about the process that I’m coming to trust.

Rapid Ideation in Isolation

In order to keep moving forward in design, I’ve learning it’s crucial to articulate your thoughts in a way that can provoke feedback.

More and more I’m finding that I work best when I have a chance to sketch out a few ideas (like above) and provoke myself before I consider those ideas with other people.


Theory Testing

It can be tempting to spend far too long designing and refining in isolation, insulated from critical feedback because your design “isn’t done”.  In order to meaningfully test a design idea, you need to approach the process with intention.

Early on, I built a crude prototype of the physical size and feel of the thermostat to be use in conjunction with paper prototypes.  Then I gave people goals to test out and and tried to put myself in their head-space as they were interacting with the design.

In this course we employed “think aloud testing” to prompt the testers to speak what they were thinking out loud in a stream of consciousness to help facilitate our own understanding of their actions (or inactions).  I took this process and had fun with it, probably due to my background in teaching and tutoring.

After a tester completes all the paper prototype scenarios, they fill our a questionarre that is scored and turned into a System Useability Score on a scale from 0-100.

I found the SUS scores to be far less useful than the think aloud testing, but I think they do represent a meaningful verification of the progression of a design project for people that are completely outside the process.  Over the course of my 4 weeks of testing, my design improved from an average SUS score in the low 80s to an average SUS score of 92 on my final paper prototypes.


If you’ve been following along with me throughout this process, you know that last time, I was looking forward to building a digital prototype of my ideas.

I’m happy to report you can check out my digital prototype online now.  Although there are still a few minor issues with how the animations are timed, I’m happy with how the digital prototype conveys a much more complete vision of how each interaction should look an feel.  It’s clear to me that in a collaborative environment, it would be extremely valuable to bring design ideas toward this fidelity before handing them off to ensure the ideas are well communicated.

Lasting Takeaways

I think this project was important for my progression toward an interaction designer who can work autonomously.  Matt Franks did a masterful job of offering feedback at each stage that pushed me forward but forced me to find my own way.  As a result, I feel very confident in the process because it’s one that I helped shape myself.  Now, as a class, we can stop thinking about thermostats and apply rapid ideation and creative problem solving to problems that are worth solving.