ResearchKit brings a hardware based solution for what amounts to crowd sourced data aggregation using the iDevice sensor suite as an open source platform that allows the contributor to have control of which data they want to share. With Apple excluding themselves from the view of data, privacy and security is maintained (at least from the collection point to the aggregation point)
The technical overview currently describes medical activity, but it doesn’t specifically say it can’t be used for other types of research. It comes with a survey module, one for consent, and then the active tasks modules. Not all of the sensors and output methods are utilized in the initial deployment. Only the mic, the gyro and accelerometer, and the digitizer are deployed. So that gives room for developers to possibly create tests and surveys based on the FFC, the speakers, the atmospheric pressure sensor the vibrator and others might be able to be used for research.
For example, if someone could develop good tracking recognition software, you could deploy an eye test. Aggregate enough data and you could start to trend eyes..the pupil, the iris, condition of the vascular network with certain health issues. I’m not a medical expert by any means, I’m just throwing out ideas.
Let’s say you wanted to see if you wanted to see how altitude affects acuity of hearing range. You could deploy the barometer, have the speaker throw out sounds at different frequencies and poll the user via survey if they can hear certain frequencies. Then GPS tag lat/long on the map to you could place an overlay showing which regions of the would might have commonality in their hearing…if there is any.
Let’s take for example a current cultural contrast: smoking weed. If it were possible to develop field test to confirm if someone is stoned using the FFC to verify the condition of the eyes, a cognitive skills test using the digitizer and a poll, and then a test that utilizes the gyro and the accelerometer for a physical movement exam you could have a pool of people who could be classed as stoned or impaired or whatever and the exclusion would be people who are not stoned. Then you could create a survey or a test to see how marijuana affect an intellectual examination or some kind of test battery. If you put a clock on time question was asked to time response input was given, you can discern if weed has cognitive impairment on reaction of cognitive performance. You could trend level of intellect on pot smokers vs non. You could ask pot smokers to take an exam while high, and one while not high, verify they are not impaired, and see if weed has an impact on that. If you could partner up with the devs who built ‘Elevate’ (one of my favorites) you could deploy the tests they built. Listening capability, reading, math etc.
What is really exciting, if when/if this data (in its non implicating and private state of course) will be available to the public. Where overlays and charts can be created either in an app or on the web…between all of this data by choosing attributes or tags. You are capturing age, sex, ethnicity, political affiliation, favorite color…the possibilities are endless. If it opens up to ‘collect anything about anyone, any time for any reason) you could trend how many stoned female democrats with blue eyes like macaroni and cheese.
Of course, all of this is Pandora’s box. Security is meant to be circumvented. Apple has done a great job staying ahead of everyone…even the Government. My only concern would be someone, or some agency…or even a foreign government compromising the data is it is being transmitted or perhaps while still on the device. That would mean the data has a name (theoretically). We know the NSA and the CIA have done this on other platforms. That would screw it up for everyone, because people would stop using it.
In this program, there exists the propensity to learn a lot about one another and ourselves. The opportunity to change lives. The opportunity to save lives…perhaps even our own.
Let’s hope Apple can keep up on top of things, so some jackass doesn’t screw this up.