1 September 2005
Wireless as a hot topic
Process and asset management see wireless innovation through new lens.
By Gabe Sierra and Bob Karschnia
Manufacturers in the process industry are investigating the viability of deploying wireless technology in their facilities.
A number of automation end-users have recently begun implementing some wireless solutions in their facilities to solve 'nuisance' monitoring problems they currently have in their plants. Conversely, basic 900MHz radio technology specifically designed for integration into Remote Operations Controllers and SCADA systems, for example, has been widely used in Oil & Gas field automation for more than 15 to 20 years. Even after some 20 years of service, the adoption rate for in-plant wireless solutions is still negligible at best. This phenomenon is probably due to the lack of industry-accepted open standards, inadequate security, and insufficient scalability/flexibility of traditional wireless solutions.
Given the long-term implications of deploying new technologies, these issues don't exactly give plant managers and engineers the confidence required to broadly deploy wireless technology into their facilities today. However, the leaders in the process industry realize these are not indefinite issues. Therefore, many are already beginning to develop advanced asset management strategies enabled by wireless innovation in order to meet their business objectives of increasing overall operational performance and/or enabling a more successful project execution.
Meanwhile, others in the process industry face the added complication of having assets, which are all over the map, such as remote customer bulk-inventory tanks. This is not a trivial problem since logistics managers fret constantly over the conflicting requirements of preventing customer tanks from running empty, which could lead to a process shutdown at a customer site, while at the same time trying to significantly reduce their distribution costs.
Effective logistics planning is not an easy thing to do well. However, several leading industrial gases, chemical, and specialty chemical companies are driving toward wireless-enabled, Web-based inventory monitoring solutions, which allow them to accomplish these critical business objectives.
In these industries, customer service and cost containment are vital to creating and maintaining a significant competitive advantage.
With such wide-ranging user needs, how can truly innovative and reliable wireless solutions develop to solve so many of the high stakes operational problems encountered throughout the industry? First, one must acknowledge there is no "magic bullet" (technology or solution) that can individually solve the myriad of in-plant and remote asset problems. That's the bad news. The good news is these end-user problems are quite solvable if one simply focuses on the root cause problem and match that need with the appropriate long-term viable solution, which will help solve that particular problem.
With the prospect of being able to solve virtually any process or asset monitoring application, many people throughout the process industry view wireless technology as an exciting new innovation path for addressing issues deemed cost prohibitive, not technically feasible, or lacking in 'device-to-host' dependability. This excitement in the asset management community is well justified with the expectation that knowing more about what is going on within the process, physical assets, and overall operations will lead to a safer and more profitable enterprise.
Technologies drive solutions
Although there are numerous wireless technologies in existence, only a few specifically target the process industries. Why? Well, the process by which wireless signals propagate through the air, the amount of data carried (bandwidth), immunity to RFI/EMI interferences, susceptibility to physical obstructions, scalability, reliability, and obviously cost vary across wireless network technologies. Therefore, many users are interested in making sense of which technologies may be applicable to them depending on the application. More to the point, they are interested in which technologies will drive solutions that leverage open standards. This is because widespread deployment will have long-term implications in the plant or asset base.
Timely information in order
For monitoring applications of geographically dispersed assets, which do not have a local host such as a DCS or RTU for data collection, Web-based monitoring solutions incorporating GSM/GPRS, or satellite communications are typically the first technologies considered. Because of the much higher costs, satellite is typically reserved for specialized low-required bandwidth applications in isolated areas where cellular coverage may not be currently available (e.g. marine vessel tracking).
Advanced Web-enabled monitoring solutions allow the combination of reliable intelligent field devices, dependable wireless capabilities with advanced power management, and a user friendly Web user interface to formulate a complete asset management solution.
For example, in a Vendor Managed Inventory application, a GSM/GPRS enabled multi-variable level measurement device will transmit monitoring data on a periodic basis (e.g. several times per hour, once per day, etc.) via an existing telecommunications network to a central server where the information is accessible using a standard internet browser. Flexible systems also allow the collected data to link to a user's proprietary database.
Some integrated solutions with suitable data logging, advanced power, and communications management, as well as measurement best practices, are capabilities, which can extend monitoring enhancements into other types of applications as well. Examples include paper chart recorder replacements, analytical monitoring for reservoir water quality, and pipeline block valve position for auditing purposes.
In these applications, asset managers require dependable access to timely information in order to establish and maintain excellent situational awareness of their remote assets. This allows the logistics planners, for example, to prioritize their distribution routes in order to make deliveries only when necessary and/or efficiently respond to quick ship requests from their customers. Another example could include giving gas distributors with regional pipeline distribution networks timely visibility into line pressure condition anomalies.
900MHz radios preferred
These technologies, if properly deployed, offer managers some new tools in optimizing their plant operations by breaking the cost barriers associated with collecting more asset performance information. Some applications could be as simple as adding cost effective measurement monitoring points to eliminate manual collection of field data, thereby improving the operator labor productivity. Alternatively, in more sophisticated applications with a centralized asset management interface, wireless innovations will enable users to extract full diagnostics data and predictive intelligence from the devices, which will then automatically notify the appropriate personnel of the precise problem before a process or asset problem can threaten a costly unit or plant shutdown. Understanding these technologies will provide insights into where they might deploy to accomplish these objectives.
A wireless system has the following components: wireless node, gateway, and host. A node refers to a wireless device installed in the field. A gateway aggregates wireless devices into an access point, which can then integrate into a host system such as a DCS, Historian, or other interface—typically via Modbus, Ethernet, OPC, or other digital protocol.
Self-organizing networks: This is an emerging yet very promising wireless technology for the process industry because it may finally enable a wireless device level platform from which open standards can effectively develop. The HART Communication Foundation Working Group is focusing on developing the Wireless HART standard whereby self-organizing networks will probably play a vital role. Since this technology was originally for commercial applications with vastly different requirements, the development efforts needed to transition into industrial applications, while still exceeding user expectations, is proceeding rather quickly. However, it will take some time before a final standard is established.
The basic concept of self-organizing networks is each individual wireless device also acts as a router for other nearby devices, thereby eliminating the need for costly site surveys. If one device cannot communicate back to the gateway because of distance or obstructions, the message will transfer to another wireless device, and so on, until a reliable path to the host is established. This is a dynamically self-managed network capability, which can perform automatic corrective actions such as auto-recognizing authorized new devices into the network, re-routing communications when needed, and optimizing the overall network performance among other things. These networks, therefore, literally heal themselves if something goes wrong with one or more devices. Ease of use is also an important consideration factor; hence, all these sophisticated tasks transpire without the need for any human intervention or additional software.
The prospect of being able to easily specify and install such devices without worrying whether or not it will work due to line-of-site issues, sacrificing device performance, ignoring device intelligence, or propagating proprietary solutions with weak security is very enticing to many users. It also holds the advantage of being more energy efficient than traditional point-to-point wireless devices because of its ability to transmit at much lower radio power requirements.
Instead of having to "scream" from the device to the gateway, self-organizing network devices can "whisper" from device to device until it gets to where it needs to go. If a direct link to the gateway cannot lock on, then the communication across one or two hops will typically solve the line of site problem, although networks will be quite capable of efficiently managing many hops. Moreover, with advanced energy management capabilities and inherently lower power wireless protocols, devices will have a longer battery life or benefit from unique energy harvesting techniques.
Self-organizing network technology will require a paradigm shift in thinking when it comes to how asset optimization can be cost effectively achieved, particularly since deploying more of these network devices will actually increase overall communication reliability, as multiple pathways will arise regardless of topology. This is counter-intuitive but nevertheless a highly valued unique attribute.
The three most common topologies are star (point-to-point), mesh (point-to-multipoint), and cluster trees, which are a hybrid of star and mesh. These solutions typically operate in the 868MHz, 900MHz, or 2.4GHz frequency bands depending on the world area requirements.
There are many different terms to describe this innovation in wireless technology. These terms (Motes, Mesh, Zigbee, and IEEE 802.15.4) serve interchangeably in the industry. Although generally accepted, there are some subtle but very important differences between those terms and self-organizing networks.
Motes strictly refer only to the actual wireless field devices/routers. Mesh is a reference to a specific topology made possible by this technology. Zigbee is the consortium, which developed a standard used for commercial (building automation, etc) applications. The IEEE 802.15.4 standard specifically refers to the radio physical layer only. The term Self Organizing Network describes the full embodiment of capabilities, technology, and industrial application of this technology.
900MHz radios and 802.11b/g wireless Ethernet: Does one compete with the other? Are they to work in conjunction with each other? Where and how are these technologies best put to use in the process industry? These are typical straightforward questions but nevertheless important ones to ask. We should consider these two technologies for use at the gateway level of a network architecture since devices do not carry these yet. Wireless Ethernet devices would be difficult to implement due to high power consumption and the market acceptance of 900MHz wireless transmitters is still well below 1% despite its long-standing availability.
In the oil and gas (O&G) field segment of the market, high power 900MHz radios have a long and successful record of accomplishment, which is not likely to change. What will change, however, is the type of data which will be added into the long distance (up to 20 miles is not unusual) communication links. As some of the gas field assets begin to mature, we should expect them to begin capturing more of the device intelligence in order to increase the overall productivity of the under-to-moderately performing well sites. In some cases, this advanced capability is deploying into new fields as well. The I/O count for these applications are typically pretty low, therefore, adding diagnostics data will not significantly challenge the bandwidth limitations of the frequency band.
Like many cities across the U.S. and abroad implementing wireless everywhere, so too are many North American enterprises including educational campuses, military installations, and industrial process plants. Many of the more innovative users are cost-effectively deploying reliable wireless Ethernet access points throughout the plant so no matter where someone is located in the facility they can securely access the plant network. This makes for some interesting use-cases for wireless handhelds in the future.
This phenomenon is significant for a several reasons. First, wireless Ethernet provides sufficient bandwidth needed to concentrate advanced monitoring data into a relatively low number of cost effective access points throughout the plant. Second, contrary to some popular beliefs, 802.11, when implemented correctly, is actually a very secure and robust open standard. Third, it is a well-understood technology in wide use within many manufacturing and business operations. Lastly, it provides the most flexible and easiest paths for data integration into DCS, PC, OPC, or other hosts.
Both of these technologies can (and already are being) be deployed into plants today. However, due to data integration at the host, required bandwidth, reduction of RF (radio frequency) noise, and enabling increased functionality, wireless Ethernet would be the preferred platform for concentrating field device applications. Because of the extreme 10-20 mile distances and relatively low bandwidth requirements, 900MHz radios will likely continue as the preferred technology for integration into SCADA systems.
What about security?
Security is by far the most frequently asked question regarding wireless technology for in-plant applications. For those concerned about the protection of wireless systems (some users are not), there are four components to an effective security strategy. State-of-the-art techniques that rely, again, on open standards are the recommended best practice. It is fact that proprietary security systems are the easiest to crack.
Encryption: This entails taking data and turning it into data packets that do not make sense to anything except a receiver that has the security key to decode the message. It is best practice to have this capability at the actual device.
Authentication/Verification: This measure ensures only valid devices trying to communicate on the network can actually do so. This prevents rogue devices or hackers from trying to gain entry through an access point.
Anti-Interference: All this does is increase the likelihood a message can transmit some distance away even with moderate obstructions and/or RFI/EMI signal interferences. The most common methods accomplishing this are Frequency Hopping Spread Spectrum (FHSS) and Direct Sequence Spread Spectrum (DSSS). Simply put, signals transmit over various frequencies, and if there is a block at one frequency, another will carry the message through. There is a common misconception that FHSS and DSSS alone can provide adequate security. For clarity, FHSS and DSSS is not the same thing as security.
Key management: Poor key (codes) management can eventually defeat any of the above security measures if a disgruntled former employee, for example, maintains access to static keys and decides to cause harm. This would be akin to losing your bankcard with your password written on the back. Obviously, neither situation is desirable, so appropriate procedures and methodology should play.
Implementing all wireless security measures properly will result in a safe and secure network that can ultimately deliver highly value to the user. Therefore, users considering industrial wireless solutions should collaborate with automation leaders who are committed to providing all aspects of secure and reliable network architecture.
Wireless innovation is creating an exciting era in process and asset management by encouraging all in the industry to see things through a new lens. Some of these solutions are already deploying in new ways for O&G, municipal water, in-plant, vendor managed inventory, and other applications. Furthermore, Wireless HART and other open industry standards are beginning to emerge, which will lead to solutions that solve many high-stakes asset management problems that are otherwise cost-prohibitive or technically challenging. W
Behind the byline
Gabe Sierra (firstname.lastname@example.org) is the wireless marketing manager for Emerson Process Management. Bob Karschnia (email@example.com) is director of technology with Emerson. They both hold engineering degrees, and they are ISA members.