At Root Technologies, we’re building a wall outlet attachment that interfaces with your existing plug-in window AC units from your phone to manage your home's climate control for you while saving time, money and the environment.

My Role

I led the interaction, web, and visual design. I conducted market research and user interviews to better understand how to best solve the problem for our prospective users.

The Challenge

Creating a Nest for plug-in window AC's at an affordable price point. Designing a hardware and software solution to help individuals save money and the environment.

Identifying the problem

Over 120 billion dollars is spent on energy used to heat or cool homes each year. And of this $120 billion, over 61% is wasted.

Although some of that waste results from factors such as poor insulation in homes and older systems, much of it actually comes from simply using our appliances inefficiently, such as forgetting to turn them off when we leave the home and setting the temperature unnecessarily high or low. Now think back to your last electricity bill.

How much did you spend on heating and cooling? Now imagine that nearly two thirds of that was money down the drain. This is a huge problem, and many companies have already offered solutions such as programmable and smart thermostats, which learn your behavior to cut out unnecessary waste.

Unfortunately, almost every solution on the market is only compatible with homes that use central heating and cooling, ignoring the 25.9 million households living in older and more urban neighborhoods in the United States.

We also performed user research to a) gain market validation and gauge interest and b) understand the user better, which is especially important when considering the design of the interface. In our user research, we asked many probing question to understand what users consider as friction points, and how they overcome these friction points (if at all). We went out to various parts of Philadelphia and interviewed 30 individuals selected at random, allowing us to anaylze the data, aggregate it, and form a user persona.

Our Findings

People are lazy and view manually turning an AC on and off as a friction point.

People leave their window AC's running all day during the Summer, so when they come home it's cool, but this results in higher energy bills.

People in Philadelphia live in older town-houses or apartment complexes, often with multiple window AC units.

User Persona

Using the collected data, we developed a user persona of an individual most likely to use our products.

It was extremely eye opening to see the reactions to the product we described; almost everyone we interviewed thought it was a great idea to help monitor and control energy usage without being at the expense of convenience or price. It seemed like our product was a great match for our targeted market.

Putting together potential solutions

Based on what we found in our market research, our target market primarily seeks convenience as the primary goal, but not at the sacrifice of paying for a high price point.

Working with a team of two electrical engineers and three software developers, we discussed the possibilities and limitations of what could be specifically created on both the software and hardware sides. I began by sketching out potential user flows to gain a better understanding of how our users might address their own pain points. There were many, many ideas that we came up with, but we narrowed them down based on a) feasability of engineering and b) if the features are core to solving real user needs.

User Flow

I worked in tandem with the engineers to develop the userflow given the software and hardware constraints. We focused on making it as easy as possible for people to sign up, given that based on our market research, people who shy away from smart devices are turned off by the perceived complexities/complications. Furthermore, developing a userflow allowed us to explore the full structure and information hierarchy of the app.

Wireframe Sketches

Building off of the user flow, I wireframed the flow to see how each individual interface element might be represented within the structure of each screen. It was also a great opportunity to further identify how features are represented, and how the information architecture is perceived.

Low Level Design Decisions

Pair with new or existing root?

In the sign up process, we accounted for the use case of pairing with a new root or existing root. Many individuals live with roommates or a family that might also want to be able to control their window AC temperature remotely as well.

If an individual decides to pair with an existing root, the phone will scan for the root then promptly send a notification to it's original owner to grant permission to use root.

Home Screen

From our user research, we also learned that many individuals owned multiple window AC units. As a result, the left screen below provides an option for users to add multiple roots. We believed the best way to organize this were vertical, elongated containers to best imitate the shape of the root itself.

For the screen on the right, allowing users to simply turn their window AC's on and off from their phones wouldn't fully turn their window AC's into a smart device. Using the built-in thermostat in the root, we can allow users to set the exact temperature of the room. We also provide an ETA for how long it will take, which provides users with immediate feedback of their action. By doing so, we turn a $80 window AC into a $250 Nest, which based on our research, does a great job in addressing users pain points of manually setting the temperature.

Gamifying Savings
By choosing a tab styled menu bar, we display a user's potential savings as efficiently as possible, because this is probably the first need our product helps solve.

Ultimately, the savings the root provides is one of the primary reasons users buy it, so we wanted users to be able to easily get an idea of how much their saving. As their savings go up, they likely become more and more excited about using the root.


We wanted to use machine learning to help control the climate (like the Nest thermostat) based on an individual's patterns to provide convenience for users. However, we also wanted to provide users with an option of manually setting their own schedules.

The design of the scheduling system was relatively difficult. Our goal was to provide users with an intuitive way of scheduling different blocks. We decided to have users use the drag interaction to set a block of time because they found this to be the most efficient interaction (once learned).

Smart Notifications

When users leave their home Wi-Fi range, we interpret this as them leaving their home/apartment. If they leave their window AC on, we send a notification to them to remind them to turn it off. We we're initially worried that sending too many notifications to users would get annoying. In order to get around this potential frustration, we integrated ML so that after the first week or two, Root begins learning your patterns and automates this process for you. In addition, as mentioned above, users can simply set a schedule for their root or turn notifications off (in settings).

Menu Design

After testing this out on potential users via InVision, users told us that they found it more valuable to have a tab style menu as opposed to a toggle menu for a variety of reasons. Firstly, users perceived this as a friction point to access core functionalities-two steps to view savings instead of one simple tap. Furthermore, we want to encourage users to view their savings as a form of gamification, so hiding it behind a toggle menu would likely be disadvantageous for us.

Lessons Learned

Ultimately, the team disbanded after the summer, and we were never able to bring this product into reality. This has definitely been a great learning experience. Working with a hardware startup is a new experience for me, and while it was challenging working directly with hardware and software developers, it was also extremely eye opening to learn about the technology itself, which is definitely something I wouldn't have learned otherwise. Given the capital constraints of a hardware startup, however, we're currently in an effort to raise capital (about $80k) to be able to manufacture our final product.

In terms of the design process, I would have ideally liked to have spent more time focusing on wireframes.

Looking back at it, the development of the wireframes felt someone rushed and could have definitely been considered more. For example, we missed a few different unlikely but possible use cases initially (i.e. root owners signing out), which we went back to correct therafter.

From a branding stand point, it was a challenge to design the brand ground up. How do we set ourselves apart from every other green tech start up? How do we "wow" our users while bringing simplicity into their lives? The website brought many of those questions into play, and we iterated through many different designs and copy styles to achieve this.

And finally, while we never officially launched the product, our website, (designed by me and developed by my friend Ian Sibner), we we're able to get over 500 sign ups.