Thoughts from the IDEO CoLab Makeathon
Prototyping an AR nutrition app for diabetics in < 5 hours
I had the pleasure of participating in IDEO’s most recent CoLab makeathon a few weeks ago in Boston. Though only a day long, the event brought together designers, developers, entrepreneurs, domain-specialists and many others with an interest in design thinking and making.
The CoLab is IDEO’s future-oriented lab. Its mandate consists in examining the intersection between emerging technologies — for 2017 that means machine learning, augmented reality (AR), virtual reality (VR), blockchain, and internet of things — and five core needs: mobility, work, health, money, and food. For the makeathon, each participant was assigned to a group of four and given a sheet with two categories: a particular technology and domain. As for myself, I was paired with a business student from MIT, a designer from Parsons, and another designer from Cornell. We got a pretty interesting pairing: health + AR.
Iterating on “Diabeats”
Based on our past experiences and interests, we quickly decided as a team to focus on nutrition-related needs for diabetics. For the first hour or so we scoped down our problem space, did some quick research online, and even got a chance to speak with a diabetic for about 15 minutes. Given we only had 4 hours to prototype a design, we narrowed it down further to focus specifically on adults with recently diagnosed type II diabetes — a group we realized tended to radically change their nutritional habits once diagnosed.
After brainstorming and sketching dozens of ideas on post-its, we chose to start prototyping an augmented reality app that could help new diabetics browse a restaurant menu and assess the impact of different dishes and foods on their glucose level. We coined our project “Diabeats”
Prototype #1: nutritional information in AR for a menu
To experience what it would be like to obtain nutritional information while glancing at a menu we quickly drew one up and created strips of paper with different types of information about the menu items: everything ranging from specific data points (i.e. number of calories, glycemic index) to abstracted data (i.e. number of stars as a representation of sugar level) to personalized suggestions (i.e. “Eating this will take you over your suggested daily sugar intake”). These strips represented the AR content we hoped to overlay over the menu. Though crude and simple, paper prototyping remains one of the fastest ways to test out design ideas — even if you have to pretend the paper represents digital content.
Using other folks in the event as test users, we confirmed that personalized information and data presented in context was more meaningful than raw data. Numbers meant very little, despite our noting that many menus nowadays contain calorie labels. People also wondered how to activate and disable the system. How would you turn the AR information on/off while glancing at the menu?
Prototype #2: The group dining experience
Next we bodystormed whether some kind of collaborative AR-based suggestion system for friends might be useful if one person were diabetic but they all wanted to share a meal. In our guerrilla research we discovered that some diabetics value being private about their condition, trying to minimize the disruptions it could have on their normal activities. Could we create AR menu overlays to help a group of friends choose a few dishes to share while keeping in mind the diabetic person’s food requirements?
Though we didn’t feel we truly tested this idea out, we still discovered a few useful details. For example, transparency of information doesn’t always cause a change in behavior; one tester felt it was easy to ignore the system’s suggestions if he had a craving for a certain dish. This helped us consider whether a system that suggested modifications to a dish might be more useful than one that flat-out rejected a few options on the menu.
Prototype #3: Expected glycemic level indicator
The previous exploration led us to consider ingredient-based menus where diners create their own dishes by selecting from a list of ingredient options. Perhaps an AR interface could reflect the health impact of different ingredient combinations, helping someone mix and match ingredients while dynamically viewing the expected impact on total blood glucose level. With this prototype, we confirmed that food options for diabetics did not have to reflect binary choices (i.e. safe to eat or not), but could help diners moderate what they eat throughout the day. Systems that give more control and freedom of choice to users began to feel more appealing.
Prototype #4: Putting it all together + creating real AR content
As a way to combine all our previous explorations and test out some native AR content, we combined a few elements from the previous prototypes and created some minimal UI mockups. Just like the rest of our prototypes, these were done very quickly! All in all we had about 4 hours for the actual group work/making session. Using Unity, Vuforia markers, a webcam, and a menu printout, we simulated how the mockups would look as native AR content. It did not look polished, but we were able to communicate the idea directly with this prototype.
By far the best part of the experience was working with my teammates. They were a smart, funny, incredibly talented bunch. Our skills were well-balanced (kudos to the IDEO organizers) and spanned visual design, business acumen, prototyping chops, design methods, and knowledge of the healthcare industry. But most of all they knew how to listen, provide feedback, and incorporate what others were saying into their own design proposals.
It was also great to see how much we got done in just a few hours. With limited time, it’s amazing how creative and resourceful we got in order to build our prototypes. And yet, we learned so much from these simple artifacts, they were crucial in solidifying our thinking and helping us ask the right questions. I’m grateful to the IDEO organizers for a good experience, and doubly grateful I was selected for the CoLab fellowship program! Will be in Boston soon for another round of making.
