April 2009

Special Section: Wireless

Get your feet wet

Industrial users can get started with wireless implementation by following a few guidelines

FAST FORWARD

  • Determine plant viability for wireless, first step in premier implementation.
  • Determine scope, and do a site survey.
  • Do not forget about security measures.
 
By Scott W. Sommer
Watch1

Talk of wireless instrumentation and wireless applications is all around us. But, practically speaking, how do we, as engineering professionals, break the ice with our operating units on installing wireless devices in the plant? Once we have determined wireless is a viable option for an application, how can we help our organizations take the first step into the world of wireless with minimum risk? First do the proper preparation and planning. Properly tuning a wireless network can be tedious and take time, so planning for both is essential.

What to consider

Options abound in the wireless instrumentation world just as the wired world. Vendors offer various wireless transmitters, wireless data concentration devices, and other radio-based devices. Different vendors offer wireless communication using different radio sets and communication protocols. Therefore, in developing a wireless network, the engineer needs to determine the best way to "wire" the wireless devices. The main differentiation between wired and wireless systems is wireless systems share the same data transmission medium with other devices.

With wired systems, you can add instruments by adding more wires. The limitation is the available I/O terminations at the control system. With wireless systems, the limitation is the bandwidth available for all instrument connections. This bandwidth is not unlimited and will eventually be exhausted in any wireless network for a given frequency band. Unfortunately, determining where you reach this limitation in a wireless network is not easy, or even possible, until you reach that saturation point.

You will need to complete certain activities before implementing a wireless instrumentation network in order to select the proper hardware devices, to insure reliable and secure data transmission, and provide the required throughput and data handling:

  • Determine the features, data throughput, and update rates required.
  • Complete a wireless site survey to develop the correct topology and architecture.
  • Construct the proper suite of applications required to process the data collected from wireless devices.
  • Develop a security and failure recovery methodology in case of transmission errors, power failure, etc.

Physically, the essential requirements of any wireless instrument network are simple:

  • Instrument or control device with a transmitting radio (wireless instrument)
  • Power source for the wireless instrument
  • Receiver (access point or gateway) capable of  receiving communication from the wireless  instrument or device
  • Power source for the access point or gateway

In order for the wireless instrument and the access point/gateway to communicate, consider these wireless characteristics:

  • Physical characteristics of the space in which your wireless network will operate, including fixed tanks, pipes, etc., building materials, competing electromagnetic transmitting devices, and other interferences.
  • Communication protocol following published and emerging wireless standards and radios [WiFi (IEEE 802.11g), Zigbee (IEEE 802.15.4), Bluetooth (IEEE 802.15.1), ISA100.11a (IEEE 802.15.4), etc.]
  • Security features, which limit access to the devices and networks to authorized users
  • Applications capable of concentrating the data received from wireless devices and storing the results

Getting started

The best way to introduce wireless technology to an industrial plant is to do it slowly, using one process area, one specific piece of equipment, or one specific control loop as a starting point.

feet2

The first step toward a successful wireless implementation is to decide which instruments should be on the wireless network. Whether the initial implementation is a single loop or an entire plant process area, the several steps are required. But first, it will help you decide which instruments to include on the wireless network if you understand the major topologies wireless instrument networks use.

The simplest form of wireless instrument network is the point-to-point configuration. In this configuration, one radio at the wireless process transmitter pairs with the access point radio. This is usually the lowest cost option and one that works well for isolated process transmitters, such as the pressure at a remote location along a gas transmission line or a tank level at a loading dock.

In the expanded point-to-point topology, a single radio can concentrate the data from multiple transmitters or I/O points for transmission to the control system. This configuration would work well for integrating the multiple transmitters, valves, and discrete I/O from a portable tank or OEM equipment package. This configuration is limited to a small number of devices (10-15) per segment.

In a multi-point configuration, multiple individual wireless devices are controlled from a central point in the control system. This may be the preferred method for communicating with room temperature monitoring points back to a building management system or for interfacing tank levels in a tank farm to the control system. This is also how multiple computers in a home or office wireless network communicate with a host or server computer. This configuration can handle large numbers of wireless devices, and is limited only by the gateway and the type of radio and wireless communication protocol.

A more advanced wireless instrument topology is the meshed instrument network. Each instrument within range of another instrument has the ability to transmit and receive data and pass on data from one instrument to another, until the data is transmitted to the host (control system gateway). Every instrument radio module is on a common clock, and the instruments have the ability to turn on and off to transmit or receive data. The instrument mesh is self-configuring and self-healing, adjusting the pathway that data transmission takes in case of the failure of one instrument or in case of RF interference. If one device drops off the network, the others will report the failure to the host. Although this topology depends on battery-powered instruments, the required power consumption is low.  This topology is well-suited for instrument networks of any size, with the control system gateway and its associated radio and communication protocol being the only limitation to mesh size.

The meshed node configuration provides the next level of complexity. This configuration allows nodes to communicate with each other and with the control system gateway. Individual instrument transmitters do not communicate with each other, and only need to wake up to report data. This saves battery life. The nodes are typically externally powered, which allows them to be high-powered repeaters that mesh with other nodes as required. By adding a redundant gateway, this configuration can become quite robust. However, the externally powered nodes and more complex infrastructure make this type of configuration efficient for networking large numbers of wireless instruments only.

To determine whether bi-direction data flow will be required in the application (monitoring vs. loop control), you should know the requirements for data exchange rates and formats.

Take a portable process tank, with temperature and pressure indication. The portable tank also has an outlet valve and a high- and a low-level switch. The goal is to provide a wireless interface for this tank to the control system so you can roll it up next to one of 10 reactors that will supply catalyst slurry for the reaction. The operator will connect an air line to pressure-transfer material from the portable tank to the reactor. You will need to monitor all process variables, and the outlet valve will be opened and closed by command from the control system.

Wireless site survey

Performing a wireless site survey is mandatory for a successful wireless implementation. The result of a survey is a map of the required coverage area with interferences (fixed tanks, impenetrable walls, RF interferences, etc.) that will identify proper infrastructure hardware specifications. They will also identify any risk due to interferences or dead spots up front, before you make large capital expenditures.

A site survey is simply the systematic measurement and analysis of radio signals in a defined area. The site survey will determine how many access points are required, the types and gains of antennae required, and the areas prone to interference or weak signals. The site survey will help ensure an adequate bandwidth and connection quality is available throughout the area where the wireless devices will need to operate.

Understand security issues

Security is a major issue, and common concerns include:

  1. End users may disable or not activate security features.
  2. Unauthorized personnel can access a network without making a physical connection.
  3. Network can be jammed intentionally with another radio or signal generator.
  4. Network can be jammed unintentionally by other plant equipment interferences.
  5. Encryption codes can be passed maliciously to unauthorized personnel.

These are all legitimate concerns, although with proper planning, you can eliminate or minimize each risk. The main defenses in wireless networks are proper enabling of the instruments' built-in security features and proper setup of the wireless nodes and access points. Never disable security features from the installed system. Proper encryption and authentication schemes must be in place (usually selectable and configurable). More recent wireless devices have advanced provisioning schemes that require identification of a unique code in the sensor itself before being allowed to join the wireless network. Modern encryption keys and techniques make it more difficult than ever to impersonate a member of the network by an unauthorized user.

One method of insuring the integrity of the wireless network is one that most installations do not take advantage of. Most current access points, gateways, and powered nodes have the ability to attenuate the radio gain so that its span is confined to a smaller area, such as a process room or tank farm area. This makes it much harder for drive-by detection of the wireless network. For instance in the portable tank example, you can adjust the radio power from the data concentrator radio to span only the minimum distance required between the tank and the access point closest to the reactor. If possible, you should tune all radios so the wireless networks are not detectable outside of the physical security perimeter of the facility (gates, limited access areas, buildings, etc.).

It is sometimes a requirement within a certain area to limit the wireless access to specific channels in the chosen radio spectrum. A plant may have several virtual networks operating within the IEEE 802.11g spectrum. One way to segregate the networks and avoid conflicts or security issues is to assign exclusive channels to each virtual network. One virtual network could use channels 1, 4, 8, and 11, while the second network could use channels 2, 5, and 9. This may not always be practical, and there may not be adequate channels to segregate all functions.

Another useful technique is to limit the access on the wireless network to specific MAC addresses, specific IP addresses, or in the case of wireless instrumentation, specific instrument IDs. Some wireless network nodes and gateways have the added capability of detecting when a rogue device or user is attempting to access the network.

The keys to enhancing security, then, are proper encryption and authentication, limiting access to specific devices, and limiting the span of the radio signal to as small of an area as possible. In addition, the engineer should budget for the costs of providing the optimum security schema, including optional features, additional hardware, gateways or firewalls, and configuration and testing efforts. Some security features may slightly degrade the throughput or data rates for some wireless networks, so you should budget for these soft costs.

Once installing the system, you will need to fine tune the whole network.  During fine tuning, it is important to understand the normal portable interferences that may affect the wireless transmission. Fixed objects, including tanks and piping, are easy to account for, but portable tanks, operators, and other moveable items may also affect the transmission of data from the wireless sensors to the network.

Antenna type and placement is also critical for proper wireless network operation. Antenna placement and orientation can affect signal strength.

A program of routine monitoring should be in place from the time of installation to identify any issues with network connectivity, data integrity, changes in data throughput, and sensor health. The most common long-term issues to monitor for include antenna placement, wireless sensor battery strength, obstructions, and excessive data transmission errors or retries.

ABOUT THE AUTHOR

Scott W. Sommer, PE, CAP, is an automation technology manager with Jacobs Engineering Group, Inc., in Conshohocken, Penn.

RESOURCES