Industrial Automation Flirts with Wireless - The automation industry increasingly finds wireless attractive, and for several reasons.
• Running a wired network incurs significant labor and material costs, while wireless networks cost far less.
• Wireless offers connectivity for remote areas or areas not currently served by wired networks.
• Wireless controllers and I/O can manage devices and processes even in inaccessible areas, or areas where network wiring is difficult or impossible to install.
• And wireless can offer a way to provide proof-of-concept for a new project before incurring the expense of a wired network. For all these reasons, automation engineers are beginning to seriously consider wireless solutions (specifically WLAN, wireless Ethernet, or Wi-Fi) for all or part of their applications. But with all these good reasons to use wireless, several concerns remain. Among them are security, network performance and reliability, availability and cost of I/O components, and the necessity of choosing between wired and wireless solutions up front.
Wireless network security has been notoriously easy to compromise. Just stand outside an apartment building with a laptop and check the available wireless networks. How many can you access without even a password? But while personal wireless networks often remain insecure, security standards for business, industrial, and government use have been developed over the last several years and adopted by most organizations.
The earlier WEP (Wired Equivalent Privacy) security algorithm, which was found to have serious flaws, has been superseded by much stronger and more secure transmission algorithms.
Wi-Fi Protected Access (WPA), including the Temporal Key Integrity Protocol (TKIP), replaced the older WEP algorithm in 2003. The more recent WPA2, introduced in 2004, uses the even more secure Advanced Encryption Standard (AES) 802.11i algorithm.
WPA2’s AES algorithm is compliant with National Institute of Standards and Technology (NIST) FIPS 140-2, required by some government agencies and corporations. These standards can protect a robust communication system.
For secure communications, WPA2-compliant products should be used for industrial wireless implementations today.
Full Range of Wireless I/O and Reduced Number of Wireless Components
From the automation engineer’s viewpoint, a separate product line for wireless—or a subset of the normal wired product line—is difficult to work with. But being able to use the same I/O components in both wired and wireless networks would save time and money during design, implementation, and use.
• In the design phase, the engineer could specify I/O with confidence, knowing that he could use any I/O in the product line and that it would work with either network.
• During implementation, the same methods and costs for installing I/O and wiring to field devices would apply to both networks; there would be no need to retrain technicians. If communication is changed from wireless to wired at any time in the project, no additional costs would be incurred for I/O or field wiring.
• During system use, just one set of spares would need to be stocked for maintenance, even if both wireless and wired I/O were in place. In addition, wireless would be considerably more attractive if manufacturers required fewer wireless components overall. Both initial system costs and the cost for stocking spares would be lower.
Wired and Wireless Support in Controllers and I/O
For real flexibility, wireless controllers and I/O should support both wired and wireless communication, just like a laptop computer. If they can be used either wired or wirelessly—or, even better, both at the
same time—difficult network decisions won’t have to be made at the beginning of a project. If an engineer designs a project using wireless technology and then discovers a wired network would be better, he can still use the same hardware.
Adding a wireless interface to wired controllers and I/O also offers new options for segmenting networks. For example, critical I/O and controller traffic could use the wired network interface, while less critical maintenance, troubleshooting, or local HMI tasks could be done wirelessly with a similarly configured laptop computer.
As in a laptop computer, the functions available in the controller and I/O should remain the same, no matter which network is being used. Only the physical medium would be different, so all I/O features and supported protocols would be the same. Even software would not have to change. Because control and HMI programs would run on wired and wireless networks with no modification, there would be no additional cost involved for licensing, training, programming, and maintenance.
One automation manufacturer who has incorporated many of these solutions into its wireless offering is Opto 22 (opto22.com). The company recently added wireless capability to its SNAP PAC System™—an intelligent, distributed system suitable for industrial automation, remote monitoring, and data acquisition—and to its SNAP I/O™, which can also be used as remote I/O with Allen-Bradley® Logix PLCs and PC-based control systems.
• In Opto 22’s SNAP PAC System, control is distributed among programmable automation controllers (PACs) and intelligent remote I/O processors (called brains). These controllers and brains also handle all communications for the system.
• The same brains, or intelligent I/O processors, provide both local processing and communications for SNAP I/O used with A-B and PC-based systems.
Opto 22 has added wireless LAN capability to these controllers and brains without removing the communication functions that are already there.
Wired and Wireless Support
A SNAP PAC controller, for example, has two independent Ethernet interfaces (two IP addresses; two network interface cards, or NICs). These two interfaces can be used for redundant links or for segmenting networks. Adding wireless support means that the PAC now has three Ethernet interfaces (three IP addresses; three NICs)—two wired and one wireless. If desired, the wireless controller can segment a wired from a wireless network and provide a redundant link as well.
A SNAP PAC brain, in contrast, has two switched Ethernet interfaces (one IP address; one NIC). The switched interfaces allow these intelligent I/O processors to be daisy chained, if needed. Adding wireless support to the brain means adding a second IP address, so the brain can communicate through either a wired or wireless network, as required by the application.
These controllers and brains retain the same functions whether they are used wired or wirelessly.
All the standard industrial protocols supported by the wired Ethernet interface are fully supported over wireless. These protocols include ODVA’s EtherNet/IP™, Modbus®/TCP, OptoMMP, SNMP, SMTP, and FTP.
In addition, the same control and HMI development software is used regardless of physical network. Because the physical layer is abstracted, control programs running on the PAC require no modification for wireless communication.
Communications for both wired and wireless networks are set up in the same configuration software and can be done at the same time, or one network type can be added later.