In the wave of enthusiasm surrounding the IoT, medical applications are often held up as an obvious and compelling area where applications cannot fail to succeed. I beg to differ. I think there are two important reasons why almost no such applications will succeed, at least not in the way we seem to be approaching them today.
The first is based on the very wide gap between the IoT community and the medical community, compounded by a very tall barrier of regulation which any practical solution must clear. There is a “Field of Dreams” feel about this – if we (the innovators) build it, they (the doctors) will come. But that’s backwards from how innovation happens in pretty much any field. You start with a serious domain-specific problem and tentative domain-specific solutions, in this case developed and refined in a research/teaching hospital. Then you work with a partner well versed in all the regulatory hurdles (perhaps GE), to build a solution. Hardware and software come in somewhere along that path, but they don’t drive development.
There are certainly popular novelty applications with a health veneer (fitness bands for example) which are not held to these standards. But neither are they likely to be very successful outside a niche community. The information they provide is interesting, but not especially actionable, which will make it difficult to sustain interest even among early adopters and almost impossible to expand to broader markets.
The second reason is even more important. The IoT raison d’être is to generate Terabytes of data which we will distill to extract wisdom. But Eric Klein at Lemnos Labs has commented that generating and distilling is not enough. We will not have useful solutions until we also create or influence change. In the field of health management, I completely agree. Take obesity as a test problem. This is an epidemic in the West and a real solution would be immensely valuable, but surely we don’t lack for data. We have books, magazines, TV shows, apps and a $60B weight-loss market focused on this single problem – with no apparent success. The problem is not a lack of data; it is a lack of motivation. In this example, motivation must be immediate and must be designed intimately around human psychology. It should reinforce good behaviors and (within reason) provide negative feedback on bad behaviors. Again, you cannot even conceive of a useful solution until you fully understand the problem and what it will take to correct (not just describe) the problem.
While I understand the desire to push the IoT forward as fast as we can, at least in the medical domain we need to step back and focus first on problems, not solutions. I wrote a blog some time ago where I proposed a wearable solution targeting obesity which seems to violate the cardinal requirement of an IoT device – it does not connect to the Internet. Whether or not this particular solution would be successful, it illustrates that perhaps real problems are not always well aligned with how we define IoT paradigms today and force-fitting the problem to the solution is unlikely to secure a path forward. You can read the blog HERE.Share this post via: