Wireless / Fieldbus

Process Analysis Powered by Pervasive Sensors

If We Want to Take Wireless Technology to the Next Level and Identify the "Killer App," We'll Have to Stop Thinking of How to Do Wired Projects Wirelessly and Instead Replace Terminations With Connections

By Ian Verhappen

Wireless applications should not held back by terminal constraints. What additional measurements could or would you make if you were not limited by a terminal block? I believe that too many folks are limiting themselves by thinking: "What wired measurements can I replace using a basically identical wireless transmitter?" It's ironic that in the consumer market, messages can be sent to you based on your interests and where you are, yet the thought of using similar data-mining technology has not been considered by the majority of industrial users.

Some larger companies are using data mining or pattern-recognition tools in their control system or process historian to identify anomalies in the process or equipment health, and send action messages to the appropriate department; however, this is still considered leading-edge and not very common. Ironically, using pattern recognition from a field of dust-sized motes in a wireless mesh network transmitting a large range of different measurements to form a "picture" of what is actually present was what Dust Networks was striving to create when it developed the WirelessHART chipset.
The closest we've come to the concept of "pervasive sensors" is using dust or mote networks for measurement not only in industrial situations, but also in medical and other fields. Pervasive sensors must be able to go anywhere and measure anything, which requires low power and wireless connectivity.

Read "Wanted: Pervasive Sensor Network Expertise"

In the presentations from the past three Passive Wireless Sensors workshops presented by the ISA Communications Division, you'll find both active and passive RFID-based sensor developments and applications in a number of industries. Some typical industrial applications where I could see pervasive sensors being useful are, to name a few:

  • In the process, or perhaps the catalyst, in either a fixed or fluidized reactor to fully understand not just the state of the reaction, but if there are any "inactive" zones in the vessel;
  • At interfaces, especially of rag layers, to better understand the dynamics of why a fluid is not separating; and
  • Multiphase flow systems.

In conversation with your process or plant engineer, I'm sure you could also identify a list of "I wish I could measure" items.

Because pervasive networks are likely to make more complex measurements than pressure or temperature, they're apt to be used in much the same way as a process analyzer or to measure similar variables, and may require the same level of support, at least for their initial installation. Pervasive sensors are in many ways similar to the inferential sensors used to replace continuous emission measurement systems (CEMS), so combining this functionality with their novelty also makes them likely to be part of the analyzer network and supported by that team.

As a process analyzer support person, I know that if I were responsible for a pervasive network that used wireless, I would also make sure I had the infrastructure in place, including a backhaul system, so I could remotely access and monitor the system. Of course, if I have a backhaul infrastructure in place, particularly if it's dedicated to my analytical measurements, I'm also going to see how it might be used for other things, for example, connecting to the traditional larger wired analyzers or building a process analyzer maintenance network.

Facilities always need more information to better understand and control processes. Pervasive sensors and the underlying mathematical models that can be built as a result are other tools in our arsenal to make that possible, but doing so will require that we think and act differently, stretching our infrastructure to do more than one thing.

If we want to take wireless technology to the next level and identify the "killer app," we'll have to stop thinking of how to do wired projects wirelessly and instead replace terminations with connections. What connections are you now able and ready to make without any wires?