Industrial sensors cost far more to install than they should, mostly wiring and conduit, yet deliver only raw numbers. The speaker argues companies should buy cheap, durable sensors that talk on their own, then spend the savings on turning that data into something useful.
This point is often lost with people. When we talk about sensors, sort of that level 0 in the Purdue model, in the past you would always think about, I need to run conduit, I've got to tie in back at the bucket, I got to pull the wires, I got to do this, that, blah, blah, blah, right? And so all of a sudden, a $500 sensor costs me $2,500 to install. And what do I get out of that? I don't have any intelligence, right? some data. What we believe that what we'd like to be able to focus on— let's flip the tables a little bit. Let's find inexpensive, robust sensors that communicate natively to do things, right? And bring that data in, and let's spend the money, the time, the energy on how do you turn that into intelligence.
