One of the more interesting conversations I’ve had in the Industrial Internet of Things space was on the topic of sensors – in particularly hostile environments, collecting particularly difficult information. A theoretical1The facts have been changed to protect the IP example might be the need to measure operating performance for a specific type of pipe joint, used to join dissimilar materials in a pipeline deep in the ocean. The material in the pipeline is caustic, and the pressures and temperatures in deep water tend to wear away at the components of the pipe joint. So much so, in fact, that regular preventative maintenance is required to shut down the flow, dive down to the connections, and replace the pipe joint on a regular basis – to ensure the joint does not break after too much time in this hostile environment. An expensive proposition (down time, labor hours, and materials), as you might imagine.
How might data add value to this situation? Well, if I could attach sensors to measure temperatures, pressure, and material stability inside and around the pipe, I would be able to develop models that predict failure. Armed with data like this, I could arguably do less preventative maintenance work, and only spend the time and money to dive down and make the repairs when the work is required.
Sounds great – but there are some amazingly difficult challenges here. How do you “sense” material stability, temperature and pressure at 10000 feet under water? I’m not sure my Arduino board is going to survive …
A few days after that conversation, I was talking with a patent attorney familiar with this particular industrial space, trying to educate myself on the patentability of information systems based on data gathered from Things – and I told the story of my unmeasurable environment. “Well, maybe not so unmeasurable”, he said, pulling at the stack of paper on his desk. “Why, right here, I have some patent information on a method for sensing temperature and pressure in deep water environments …”. I was stunned – I thought it was a silly question, asking for an unmeasurable metric – and it turns out someone had already done it.
Or at least part of it (… hmmm, still need to figure out material stability …), but in truth, this was a reminder of a classic bit of wisdom. There are relatively few truly new ideas – just new ways to apply existing thinking. The trick in the IoT space (really, any product engineering space) is to figure out where the expertise may exist, and how to find out if an idea or concept has already been tried. It’s kind of like Googling for source code – that particular search tool points to “knowledge bases” that run a good chance of returning what you are looking for. But I don’t believe Google points to the USPTO database2Oops, oops, apparently it does – I love this job, every corner I turn there are more interesting things to learn about … or any of the many specialized, industry / market / technology specific knowledge bases that exists – inside and outside of the company you work for.
And so, your solution design process must always include imaginative, active ways to keep your eyes and ears open, getting connected with as many information sources as possible. I went to a recent Gartner Symposium and attended a breakout session on submarine technology3Again … the facts have been changed to protect the IP, and heard engineers in that room talking about sensors on the outside of the sub taking all sorts of interesting measurements.
You never know what connection will turn up the critical piece of information … so keep investing the time, keep developing your research sources!
# 12 February, 2015