Look, Ma! No wires!

Jan. 24, 2005
CONTROL magazine's Editor in Chief Walt Boyes tells us whether wireless networking in process automation applications is ready for prime time in this "special to the web" editorial for ControlGlobal.com
 By Walt Boyes, Editor in Chief

M

uch has been said about wireless networking over the past few years. Wireless networks are ubiquitous in cafes, airports, hotels, and some cities have established agencies dedicated to “wireless enabling” their entire town. A lot has been said about wireless networking in the process automation area, also. Companies have announced “hardened” wireless routers and access points, low power transmitters, small footprint devices, and some of them are even being used. Use of wireless networking in process automation appears to be lagging well behind the hype.

The truth is that a wire is still the safest, fastest way to get data from Point A, a field sensor or final control element, to Point B, the I/O of a control system.

Wires are immune to external reception by unauthorized individuals. Wires are immune to much electrical interference. Properly shielded, wires are immune to radio frequency interference. Wires are easy to troubleshoot, too. If the signal is at one end of the wire, and not at the other end, you simply follow the wire until you find the cause of signal loss. Wireless systems are a lot more problematic.

So why the interest in wireless networking in process automation? Cost is the reason. Not the cost of the network, but the cost of wire. It costs between $1500 and $3500 a foot (roughly 30 cm for metric folks) to wire a single process connection. In the bad old analog days, you could double that to close a single loop from sensor to controller to final control device. In these enlightened fieldbus times, you can still approach that number for costing a single segment of a chickenfoot bus segment.

So where is wireless taking hold? Not in process, but in asset optimization. It used to be that the thought of automated data collection on vibration and temperature, for instance, on every motor and gear reducer in the plant was unthinkably expensive. With low power dedicated transmitters and sensors, it is becoming even practical to collect that level of data. With modern asset management systems, it is practical to collect that level of data and actually use it to optimize the maintenance of those motors and gear reducers. And that pays for itself. So here, wireless is becoming a no brainer.

The fact is that if a wireless transmitter goes down, in an asset management scheme, there is little or no effect on the CMMS system.

But what if a wireless transmitter on a control loop was to go down? You can lose much more than the installed cost of hardwiring that loop in a very short time. Some loops are so critical to performance that they can shut down an entire process, costing thousands or even hundreds of thousands of dollars an hour until the loop is repaired.

So, if wireless networking is to become ubiquitous in process automation, like it has become in the home, it will have to overcome some serious obstacles.

First, the robustness of the signal needs to become as “near-bulletproof” as a simple wired connection. No mysterious signal losses, no static scramble, no failure of the carrier.

Second, the security of the signal needs to become as impenetrable as a simple wired connection. And data transmission speed losses caused by encryption need to be minimal.
Third, wireless self-diagnostics and self-repair need to become increasingly robust, so that the life expectancy of the wireless network can approach that of hard wiring.

When wireless networking provides those features, we will truly see the vision of the wireless companies: plants where every moving part is monitored, every process step is controlled, and the control system acts like a huge autonomous nervous system in an organic body, while the operators supervise and intervene only when necessary.

That’s the vision. The reality may or may not approach it.