Wow – this check-in is long overdue. In the interest of keeping up with my goals earlier this year, here's my semi-monthly summary of work, projects, etc.
First of all, let's get the inconvenient bits out of the way. Other than maintenance tasks, I made no discernable progress on pushing Fresh Comics and the Pnakotic Atlas to where I want them to end up. Life – in the form of my day job and some new commitments – predictably got in the way. I'm currently starting a mini-vacation where I'll stick around the Windy City and am working this week towards getting some solid momentum behind these projects. At the moment, I'm preparing a new Windows 8.1 virtual machine template that I'll use as the basis for new Android and Windows Phone development. I'm looking forward towards getting both apps up and running under a more modern development framework (Android Studio and Gradle) and running the ball down the field.
Now, onto the convenient bits…
The bulk of my time and attention during March and the first two weeks of April was spent getting the IntelliCare project across the finish line. This involved updating and shepherding 13 native Android apps across the finish line, as well as writing a web interface for the coaches and support staff to use to interact with participants in our field trial. This second bit involved combining Django with our home-grown event-logging infrastructure (home-grown for IRB reasons) to give coaches an almost real-time view into how study participants are using the apps. In addition, I built a chat-like interface that leverages Twilio to give coaches a window to correspond with their charges using SMS text messaging. This also includes some nifty bits that attempt to determine if the user is sending suicidal messages and alerts staff if such evidence is found.
The launch went pretty well, but was also turbo-charged when the local Chicago media decided to report on the apps. We received coverage in print, television, and radio for the IntelliCare suite:
FOX 32 Chicago: New phone apps help manage, treat depression and anxiety
Chicago Tribune: Worry Knot, iCope and a suite of IntelliCare apps from Northwestern
WBEZ Chicago: Apps offer mobile therapy
The project is now in the state where the apps are out there and my feedback pipeline has allowed me to address the major issues we encountered, so the apps (all 13) are in more of a maintenance mode, as opposed to being in active development. I have a few more tasks to complete, but after 18+ months of development, that project's in great shape. It took an enormous amount of effort and attention – same kind and order of magnitude required to launch a startup – but the long hours and Red Bull-instigated migraines were certainly worth it. The only thing that's lacking are iPhone versions of the apps, if our feedback is to be believed.
My second major accomplishment over the month of March (and early April) was going from a dabbler in the area of wearable software development to someone doing it for a specific purpose with specific aims. If that sentence is Greek to you, it basically means that I'm now in the business of writing mobile apps for watches and other wrist-based devices (but not for the Apple Watch, yet). I've been dabbling in Pebble development for months, but I finally had a client who actually needed the technology after finding that commodity devices (FitBit, Jawbone, iHealth) were not up to the task of generating decent motion data for detecting/predicting when its wearer may be likely to suffer a bipolar episode.
With Dr. Evan Goulding, I've been evaluating whether Android Wear or Pebble devices are better suited for this purposes. Both of these devices include a built-in accelerometer that samples the acceleration vector (mainly gravity) and this allows us to determine motion and other useful properties of the wearer at a given point in time. Using my Purple Robot app, I implemented some additional probes that allow the mobile phone to periodically communicate with the wearable device and retrieve stored data to transmit to a cloud server for later analysis.
At the moment, the Pebble looks to be the more solid platform for this purpose, owing to its better battery life and stability. (My Android Wear device reboots itself several times a day – I don't know if that's my fault or the device's.)
I'm excited about this development because I've not only been able to support our client, but I've also been able to productize a lot of the work and make it available to Purple Robot users. This puts us out in front of similar kinds of context-sensing toolkits and gives us a sizable advantage early in wearable technologies' adoption lifecycle.
In terms of other obligations over the past month, I've resumed some minor freelance activity and I've signed on with a talented group to produce an interactive children's book for the iPad. These break my "no new projects" rule, but the opportunities to build something interesting and work with some high-end folks were too good to pass up. My challenge now is taking the energy and attention I devoted to IntelliCare over the past couple of months and direct it to these new projects without slipping back into a comfortable state and letting that drive fade. So, while I'm nominally on vacation this week, I fully expect to put in as much time and effort into my projects and these new ones as I would during a regular work week. The goal is to exit the week with some solid momentum behind each project and be in a state where two of them are close to delivery by the time I walk back into my office at CBITs.