In the beginning, there was SCADA (Supervisory Control and Data Acquisition). Since the 1960s, engineers have been marrying radios and flow control devices. Beginning in the oil and gas pipeline industry, and quickly spreading to the water and wastewater and utilities industries, radio SCADA telemetry has become ubiquitous.
Since pipelines, water, wastewater and electrical utilities have the need to control devices whose physical locations are often many miles apart, the use of hard-wired controls as in a process plant simply was impossible.
The cost of leasing dedicated lines from the telephone utility began to be prohibitive in the late 1960s, and by the mid-1970s, early digital radio telemetry systems (with transmission algorithms borrowed from NASA) began to make it possible to install a flowmeter at a location, and monitor its signal from a control panel many miles away, and use that signal to control a final element at a completely different location, far away from the other two places.
For a long time, the cost of radios and the relatively few available frequencies made SCADA a proposition that was very limited.
Then in the 1980s, the explosion of cellular frequencies and digital communications gave radio telemetry and SCADA a new impetus. First were autodialers that used cell phones instead of landlines to call in their alarms. This quickly migrated to process plants, used for alarm functions in outlying areas like the waste treatment area, or storage locations.
In the early 1990s, it became possible to use the unlicensed 900 MHz ISM (Industrial, Scientific and Medical) band spread spectrum radios for line of sight transmission of signals. Several companies began manufacturing radio links for process and SCADA use. In the mid-1990s, companies started experimenting with IR (infrared) wireless links for calibrators and instruments in close proximity. In the late 1990s, new wireless technologies such as the 802.11x standards and Bluetooth were introduced.
Throughout the 1990s, there were multiple drivers toward developing a standard field architecture for process instruments that would use all of the power and capability of the intelligent field instruments that were being developed. Although based on the LAN (local area network) model, these fieldbus architectures incorporated numerous features specific to process automation networking. In the 21st Century, so far, there are two basic fieldbus standards that are in common use in flow and other process control, Profibus and Foundation Fieldbus. In most cases, Foundation Fieldbus is used for flow control, although not always. There are still significant interoperability issues between standards, and between manufacturers implementations of these standards.
The Growth of Wireless?
The argument that fieldbus architecture costs significantly less than standard analog control signal architecture is also being applied to wireless.
And this is not only wireless links to remote parts of the plant. Wireless is becoming a much more common method of control networking on the plant floor between machines and process systems. A 2002 report by Venture Development estimated the market in 2006 for industrial wireless networking will be as much as $752 million, with the majority of users being utilities and the petrochemical/chemical industries. This is a huge growth estimate, from a 2001 base of approximately $190 million.
The VDC report goes on to estimate that the largest uses of wireless in process control will be tank level monitoring and plant maintenance and calibration applications. It is not a great step from those to many flow control applications other than tank level.
LAN Convergence and COTS?
At the same time as the development of industrial fieldbus architectures, IT departments were developing networking for the office and enterprise environment. These LANs (Local Area Networks) and WANs (Wide Area Networks) and VPNs (Virtual Private Networks) made networking, and networking hardware and software extremely common, inexpensive, and simple to install and understand. Ethernet has become the standard for IT and even home networking, and products are being developed on this base for industrial Ethernet networking for the plant floor.
This means that for many plain vanilla network operations, even on the plant floor, COTS (commercial off the shelf) hardware and software solutions are available based on the huge expansion of LAN networking. This has already lowered the cost of plant control systems significantly. Plus, these COTS LAN components and systems are easily interoperable with the rest of the enterprise, permitting large scale Enterprise Integration at much lower cost than previously expected. The same COTS components and systems are in use for VPNs and WANs in many enterprise systems as well, and are migrating to the plant floor because of their ubiquity and low cost.
According to the AWWA, COTS systems for SCADA and telemetry are becoming commonplace in the utility industries and water and wastewater treatment.
This is not the first time that COTS systems migrated to the plant floor. It is now completely common to see standard PCs on the plant floor. Most small and medium sized control systems, even from major companies, use standard PCs instead of the dedicated processors of the great golden age of DCS. It is no surprise therefore, that enterprise networking systems should also migrate to the plant floor.
The enormous growth of handheld PDAs (Personal Digital Assistants) from Palm, Handspring, and the Pocket PCs has fueled the development of datalogging and calibration and monitoring software packages that run on PDAs and communicate with field instruments and control elements using the infrared technology built into the typical PDA. This adaptation of a COTS device to replace expensive proprietary calibrators and laptops was one of the first implementations of new wireless technologies.
Bluetooth is a wireless protocol specifically designed for instruments and devices in close proximity to one another, like appliances in a home or machines on the plant floor. It operates in the 2.x GHz ISM band, which makes it susceptible to interference from other radio transmissions such as 802.11 and portable telephones. It is slow, extremely redundant, and is therefore a very good local instrument net protocol since its range is nominally only 10 meters. Bluetooth calibrators and programmers are becoming more and more available, and it is likely that Bluetooth enabled flowmeter-to-control-valve packages will become common.
A new standard for communications technology in control is IEEE 802.15.4, and a non-profit consortium called the ZigBee Alliance is developing protocols using the physical layer of IEEE 802.15.4.
The focus of the ZigBee standard is specifically monitoring and control, and it permits monitoring over 75 meters, with over 250 nodes per network, with very small system resource usage and very low battery power usage. At least one company is already offering embedded web servers and ZigBee transceivers commercially. ZigBee is slower than Bluetooth, but goes farther, has more nodes, and is very cost effective. ZigBee networks interface easily with 802.11b networks for longer and faster data transmission.
It is likely that ZigBee networks will be useful for flow control applications on the plant floor and in remote stations like pumping plants and sewage lift stations where all the devices can be interconnected by a ZigBee network instead of hardwiring.
The IEEE 802.11x group are the most well known of the wireless networking standards. This is because 802.11b wireless networking has become commercially available and inexpensive. A wireless router and access point operating under the 802.11b standard can now be purchased for under $50. Wireless PC cards and wireless NICs (Network Interface Cards) are available for similar inexpensive costs. The other common 802.11x standards are 802.11a and 802.11g. 802.11g is a very high speed wireless connection, up to 54 Mbps, while 802.11b is limited to about 11 Mbps. While 802.11b is slower, it covers a wider area and appears to be more stable than 802.11g systems. 802.11x systems easily interface with almost all PC configurations and Operating Systems, including most versions of Windows, MAC OS 9 and X, and most Linux/Unix distributions.
Wireless Medium Challenges
The 802.11x group of standards illustrates the challenges of wireless in the plant floor environment.
As Professor Antti Kaunonen of the Tampere University of Technology, and Metso Automation, explained in a talk given at ARC Forum 2002, wireless issues include:
Unreliable in nature
– Physical environment varies in time ?
– Topology of wireless network changes?
– Connectionless, no means to guarantee uninterruptible operation
Sources of errors
– Interference with other wireless networks and systems?
– Disturbances from nature and equipment?
– Absorption of signals, reflections, multipath propagation
– Available frequencies are a scarce resource
– Increasing reliability sacrifices bandwidth
– Network coverage is not accurately limited?
– Users cannot be recognised by a physical connection?
– Intentional attacks, eavesdrop, reproduction
Need for power
All of these are true for the 802.11x group of standards. The reliability issue, coupled with the security concerns of WEP encryption and other security standards, are the critical issues for 802.11x networking in the plant environment.
As Professor Kaunonen noted in his talk, it is impossible to guarantee uninterruptible operation, because there is no physical connection. In the workplace LAN environment, or in home or home office networking, sudden loss of the network signal is survivable, if frustrating. In the plant control system environment, sudden loss of the network signal may cause a significant industrial accident.
802.11x networks interfere with Bluetooth and other ISM band transmissions, and those other applications interfere with 802.11x as well. 802.11b networks can interfere with 802.11g networks, as well, although 802.11g networks are supposed to be interoperable with 802.11b networks. In addition, interference from stray EMI, sunspots, other natural disturbances and all sorts of antenna propagation effects make it difficult to assure the reliability of wireless networks.
The other significant issue is that of network security. In my own home network, I can either be on my own network or I can be on a neighbors unsecure network, whichever I please. Since my network is secure, my neighbor will find it difficult to return the favor. It will surprise him to find that his transmissions reach my house, and it is not possible for him to limit his coverage to just his own property. However, even the best wireless security is vulnerable. WEP 128-bit encryption can be cracked, and it is becoming much more common to have hacked systems on the plant floor. Because of the inability to control the wireless networks coverage area, someone outside the physical plant property could gain access to a plants wireless network and cause damage to the plant control system and the plant itself.
Wireless Sensors and Control Elements
If you look at the amount of wire necessary to produce a manufacturing cell or process network, and you consider the time and money necessary to design and procure and assemble those wires, you can see why there is a significant market pull toward inexpensive and reliable wireless enabled sensors and actuators and final control elements.
Rotating machinery, cranes, robots and other similar machines are much easier and less costly to manufacture with wireless controls than with wired systems. Process plants can be reconfigured much less expensively and much more swiftly if the wiring doesnt have to be completely re-done just to move a reactor from one side of the plant floor to another.
Already wireless enabled level and flow instruments, temperature sensors, and even pH and ORP sensors are readily available, with many more types of sensors to come in the next few years.
The growth of wireless networks on the plant floor is easing the integration of sensors into the enterprise networks for production and inventory control and supply chain consolidation. The much lower cost of wirelessly enabling sensors is making it possible to provide data to the enterprise at higher data throughput rates and lower cost than previously envisioned.
The Future of Wireless
The mantra of wireless proponents has been: Internet+Wireless = Internet everywhere. The growth of the Internet, and wireless networking in the enterprise and home environments, has made it easy to envision a plant floor with intelligent flowmeters and other control devices with embedded web servers and wireless transceivers and the ability to communicate with themselves and with process control staff over the Internet from anywhere to anywhere.
In order for this vision to occur, wireless technologies must deal with the issues that make wireless considerably less reliable than wires. Specifically, wireless technologies must determine how to maintain high data throughput rates while increasing reliability of communications, even in noisy and difficult environments. Additionally, wireless technologies must become considerably more difficult to hack, and much more resistant to unauthorized access from both inside and outside the plant.
The market pull of wireless is strong, and we should expect to see these issues dealt with in the near future.
Walt Boyes is a is a principal in Spitzer and Boyes LLC, offering engineering, expert witness, development, marketing, and distribution consulting for manufacturing and automation companies.