In the future: More, better, cheaper sensors

By Walt Boyes, Editor in Chief

IN THE PLANT of the future, simple and inexpensive sensors are essential. We won’t have the levels of human maintenance and operations personnel we’ve had in the past. For one thing, the Baby Boomer generation is getting old and retiring, and the succeeding generations are much smaller numerically. We’ll have to fill some of those jobs with automation instead of warm bodies, because they won’t be there. We’ll be able to have something close to the “lights out” plant, but we’ll need ways to see in the dark.

The key to handling this problem is to have more sensors, monitoring more things. We’ll need sensors that we don’t even have yet and, those we do, like flow and level, need to be much less expensive so we can use more of them per plant.

"Process automation has always suffered from being a niche industry. There is little economy of scale, and there is less incentive to commoditize."

I’ve talked about inexpensive sensors in this space before. Now it is the turn of level measurement. Level measurement with differential pressure transducers is well understood and relatively inexpensive. So is the bubbler level system, whether it uses air or inert gases like nitrogen. However, both types of pressure-based systems are notorious for fouling and require maintenance. Other types of continuous level systems are more expensive, up to the most expensive of all: gamma nuclear level gauges. Non-invasive level measurements historically have cost more than pressure, floats, and other invasive and mechanical measurements. What everybody wants is good cheap level measurements. But they also want low maintenance and, especially in hazardous materials, as non-invasive a measurement as they can get.Often the impetus for really inexpensive sensors comes from outside the process automation industry. For example, the first really cheap ultrasonic level sensors were made with the sensor that Polaroid developed for their folding instant camera’s autofocus control. Now we see ultrasonic sensors, like one I saw recently from Walchem, selling for under $500. Recently I saw a noninvasive sensor from the recreational vehicle industry that is moving into tank farm and tank truck level sensing. I own a travel trailer, and the single most troublesome instrument on it is the waste tank level indicator. I was looking for something better when I came across something called SeeLevel from Garnet Technologies. They’ve taken a magnetostrictive sensor and made it flexible enough to mount on the outside of tanks. They’ve made three versions of it: one for RV tank level, one for tank trucks, and one for tank farms. The fact that it was originally developed for the RV industry makes it incredibly cheap in comparison with other devices of similar accuracy. And the fact that it was originally developed for the RV industry has a lot to do with how simple to operate and how easy it is to repair, too.Process instrumentation has always suffered from being a niche industry. Instead of making millions of sensors, like TI makes millions of calculators, sensor manufacturers make hundreds or thousands. There is little economy of scale, and there is less incentive to commoditize. To produce the high volume of sensors we’ll need to run the plants of the future, though, and do it economically enough to make the “lights out” plant practical, instrumentation companies are going to have to re-think their design criteria, just like Walchem and Garnet, and others, have done. We need to see a lot more simple, reliable, economical sensors for level and many other measurements, and we need to see them soon.