Connectivity is eating the world
The Internet of Things (IoT) extends the intelligence and connectivity of traditional computing devices to the everyday objects we own and use. Currently, interaction with IoT devices relies on discrete mobile applications for every type of smart-thing that a person uses. Network complexty grows with each new node, and with 50 billion of these smart-things projected to be configured by 2020, the friction associated with this interaction paradigm will need to be significantly decreased if manufacturers expect adoption rates to grow.
For my senior design project, I formed a team with a mission to build the first popular browser for the Internet of Things. Our challenge was to leverage existing technology including smartphones, bluetooth and wifi to allow compatibility between cross-manufacturer devices and with our browser. It had to afford intuitive device selection, effective monitoring and precise control while being robust enough to account for the heterogeneity of IoT that includes everything from baby bottles and speakers, t-shirts and parking spots, to health monitors and entire transportation grids.
In this 7 month project (08/15 - 05/16), I led a cross-disciplinary team of 8 students to both design and build the browser, as well as design, build, and modified smart devices to worked with the browser. My roles included research, interaction design, product management and user testing.
Why make it so difficult to turn on my lightbulb?
Working on the Internet of Things was a great opportunity because although it leverages existing technology, the implications - pervasive data gathering and connectivity - were uncharted territory. In order to design a browser for IoT we had to uncover the advance system interaction techniques involved. We couldn’t find anything as well documented as personal computers, smartphones or even the ever approaching, Virtual Reality. So we had the opportunity to write the book ourselves. IoT interaction techniques include:
- Discovery - Knowing which of the things around you are in fact smart and interactive
- Selection - Choosing the exact thing that I wants to interact within the universe of things around me
- Monitoring - Checking on or subscribing to status (or history) of a smart-thing
- Control - The ability to issue commands to a smart thing to change state
- Coordination - The ability to automate certain control sequences when predefined conditions are met
There is plenty of room to improve due to two main failings of the industry. The first failing is App Myopia, which is the over-reliance on discreet mobile apps for every smart-thing manufacturer. With 4-5 steps before an app can even be opened, the Device Selection Bottleneck snowballs into affecting every other technique. The second failing is a Walled Gardens where each manufacturer establishes their own proprietery network to lock consumers into their ecosystem of products. This lack of coordination completely ignores the spirit of IoT - making my life more efficient through pervasive connectivity and uncovered insights.
Building our next superpower!
The Internet of Things promises a wave of effeciency that will dwarf that of personal and mobile computers. We are in the early stages of it and we have the opportunity to lay the foundation for people will interact with the computer-enabled world around them. This is the chance to build humanities next superpower, the ability to detect, predict and improve patterns to make my life more effecient.
- Multiple Affordances for Selection: Allow users to easily select a smart-thing whether or not they can see it, whether or not they know the name of it, within or out of reach, and whether or not its unique in its environment
- Leverage the Physical Web: Google’s open-source standard for interaction relies on connecting smart-things directly to their own web app for control and makes them discoverable through bluetooth
- Monitoring and Configuration: Users should be able to organize the devices they own into saved location and easily monitor or configure devices them remotely.
Mapping the Content
Structuring the interfaces
Exploring different ways to design the main page: Users must be able to select devices with ease. For devices within reach and line of sight, we made it easy to select by including a camera view that is active as soon as the app is loaded. We also needed to list the nearby devices in order of proximity to users - based on bluetooth signal strength - for users to swipe through or search. We entertained the notion of categorizing the different types of smart things into 3-4 main groups but it seemed the categories would be far too broad and vague to be useful. From this screen users also needed to be able to access other functions of the app so I played with different navigation styles including the hamburger menu, as well as bottom and top navigation.
Tweaking the look and feel
I experimented with creating a distinct visual language for the browser, but decided to stick to the Material Design language since the ultimate goal for this browser is for it to be integrated into mobile operating systems as a utility for seemless interaction with IoT.
Does this thing even work?
As per requirement of our senior design project, we were to pick and test a performance metric that our implementation improves over alternate solutions. Our goal was to eliminate the Device-Seleciton bottleneck, so we used the GOMS (Goals Operators Methods Selection) analysis technique to verify the usability of our browser. This called for us to compare the Task Completion time and steps of the traditional methods against that of the IoT browser that we have built. On average, our solution increases user performance by 3.16 steps. This results in decreased usage time and improved task efficiency.
We've only just begun
This project spawned from self-guided research I had been doing on IoT usability. After rounds of proposals, I gained approval from UT Dallas Senior Design faculty to build a cross-disciplinary team to develop it.
I lead our team that placed fourth in the senior design presentations - the highest for any team without a coorporate sponsor at the time. The UT Dallas magazine interviewed our team and featured us in the Alumni Magazine.
I later applied for and was granted the Google IoT Technology Research Award that would fund this project through the summer of 2016.