Archive for July, 2010

DSC00222Upgrading the control system for a municipal wastewater system was going to be costly and time consuming until facilities engineers reviewed available wireless technologies. The Persigo Waste Water Treatment Plant serves the City of Grand Junction, Colo., and the surrounding communities. The facility, which cost about $28 million, was put into service on Jan. 16, 1984.

 After two decades of operation, the existing control infrastructure was in need of replacement and upgrade. Although, it originally had been designed to use the most reliable communications signaling technology available at the time, buried multi-conductor copper cables, age and constant use had heavily degraded the performance. As a result, maintaining the system became expensive.

Ed Tankersley, Lead Plant Mechanic, summed up the situation at the time: “Our wires underground were failing; therefore we were trying to get our plant-wide SCADA system running so everything in the plant would come back via wireless Ethernet. However, we had a relatively primitive communications system.”
Additionally, Tankersley realized, the entire system could be made more reliable, easier to maintain and more energy efficient if the controls for the dozens of pumps, blowers, valves and sensors were upgraded to take advantage of the latest Totally Integrated Automation (TIA) technology from Siemens Industry, Inc. The system Persigo now employs has two types of wireless data communications and a full complement of TIA-based controllers, HMIs and SCADA software.

To plan the system upgrade, Tankersley contacted Cody Lange at automation-technology distributor Crum Electric for help. “Ed’s vision was that he and his staff in the plant would have all the knowledge needed to not only integrate the system, but also to carry out maintenance into the future,” Lange recalls. “He wanted to have the knowledge to repair and improve the plant internally without having to call in external people or [reach out to third parties for] support.”

“My part,” Lange continues, “was primarily to help with the architecture and [provide information about] different types of I/O and processors available in the TIA portfolio. I also provided technical support as far as programming and Ed worked with Siemens’s Technical Support Hotline.”

The system Tankersley designed with Lange’s assistance relies on wireless networking throughout. For communication within the Persigo facility itself, where communications links span distances of a few hundred yards at most, they opted for a Siemens Scalance W Wireless Ethernet network. To communicate with control systems at the lift pumps miles away, they chose FHSS radio technology from wireless system developer FreeWave.

They simplified and upgraded the control architecture by adopting Siemens Totally Integrated Automation (TIA) technology. TIA is a networked control system philosophy that stresses harmonization and integration of all control-system components to reduce installation costs, with integrated diagnostics to make the installed system as easy to maintain as possible. The TIA system maximizes re-use of software components, and provides hardware components designed to work together seamlessly.

For example, Siemens S7-200 PLCs are used for small local-control systems, such as lift pumps, while Siemens S7-300 PLCs control larger, more complex equipment in the main facility. Siemens Simocode intelligent motor overloads were used on critical motor loads such as the three 300HP blowers, and Siemens WinCC SCADA (Supervisory Control and Data Acquisition) visualization software was used for plant monitoring and operation. WinCC also acts as an OPC server to Canary Labs Trend Link, logging any process variable every quarter second.  This enabled Tankersley to use Siemens Simatic Manager software to program the entire plant.

Huge Service Area
Robust long-range control communications are crucial in wastewater treatment facilities because these facilities generally serve a large geographic area, and their processing facilities tend to be spread out as well. Persigo’s service area currently contains 488.8 miles of sanitary sewer lines, and this increases each year with the continued growth in the Grand Junction area. In addition, three Sanitation Districts send sewage to Persigo WWTP for treatment, including: Central Grand Valley Sanitation District (78.6 miles), Orchard Mesa Sanitation District (34.3 miles ), and Fruitvale Sanitation District (29.9 miles).
As one would expect, the wastewater flow channels to the plant resemble tributaries draining a river-system watershed. The communications-network physical topology mimics the wastewater collection system’s topology. The network is a modified star configuration (in information network parlance). Of particular interest is control communications for the network’s lift stations.

Gravity drives wastewater flow through most of Persigo’s collection system, but some parts of the system require additional pressure boosts to pass obstacles, such as hills and valleys. Hence the need for lift stations, which are automated pumps providing the additional boost on an as-needed basis.
At each lift station Persigo is using a Siemens S7-200 PLC with a Freewave FGR2 RS 485 radio for communications. Some of the closer lift stations not only provide their own data back to the plant, but also repeat data from lift stations further away.

“Our farthest jump is 20 plus miles,” Tankersley says. “The radios’ maximum range depends on how much you want to spend on antennas and, of course, line of sight. They are only one watt, but I imagine you could get 30 miles in this terrain if you selected more powerful antennas.”

Advanced Technology

“One of the primary reasons they wanted to use wireless was because this is a very old plant,” Lange recalls, “and it is uneconomical to run copper wire out to all the locations. So wireless was really the preferred solution because it saved time, money, and effort, and eliminated the need to trench and get into the vaults.”
In addition, it is generally not good practice to run control and communications wiring in the same vaults as the 13,200VAC power cables. The communications signals can be disrupted by the electromagnetic fields produced by the power cables.

Tankersley says he considered a number of alternatives before deciding on this system: “At the very, very expensive end, you could run fiber, which is not practical. You could also run via satellite, which is another wireless technology, but requires monthly fees. Years ago we had a phone line to each station, but the monthly fees were killing us. You could also run cell phones at each station, but again you have a monthly fee.”

In the end, he chose to use Siemens’ Scalance Ethernet infrastructure, which is a combination of managed and unmanaged switches, and wireless and hard-wired switching. The system also includes multiple Siemens HMI panels and a WinCC Flexible runtime on a PC in the main operations building. This connects to the system’s data logging functions, and everything networks to an alarm dialing system.
One of Tankersley’s goals was to set the system up so that operators could access any process from any HMI panel in the entire plant or from the PC operator’s station in the main building. If, for example, an operator is in the digester building, he or she can look at the headworks, or the aeration basin, or vice-versa.
This is a huge advantage, particularly in the headworks building where all the sewage comes into the plant. “That’s not a place you want to spend a lot of time,” Tankersley says. To be able to look at the operations in that building from another location that is a little more pleasant was a key deciding factor. To not have to walk several hundred yards when diagnosing a plant issue, or to avoid going into an unpleasant environment to access data and control those stations, is more than convenient. It makes maintaining the plant much more efficient.

System Within a System
“Lift stations include a pump and a redundant pump,” Persigo’s Tankersley points out. Sometimes there’s a dry well for the pumps and a wet well next to it. Other times pumps are surface mounted, and then they have to vacuum pump the sewage into the pump to prime it. The bigger ones do have backup generators. Our biggest station has two 60 horsepower (HP) pumps, and the smallest stations have 2 HP pumps.”
Pump stations are automated by small, low I/O count control systems that run autonomously. Each pump station has a 24-V-powered S7-200 Siemens PLC as the controller. Each pump station also has a Freewave 900 MHz frequency-hopping spread-spectrum radio to communicate back to the main facility either directly or through repeaters (additional FHSS radios) via the Modbus communication protocol to a Siemens S7-200 at the main plant site.

The S7-200 is a data concentrator that brings everything into a second PC running Siemens’ Simatic NET OPC server. The OPC client is Win911, which is a commercial auto-dialer software that calls a phone list and relays messages to plant operators and maintenance staff for help based on alarm conditions.
Under normal circumstances each lift station runs autonomously, but they all report departures from normal operation, such as high wet-well level or power failures, back to the main facility. For example, some of the stations have backup generators. If the power fails, the generator starts up automatically. When the backup generator successfully starts, the power fail condition clears, but the original power-failure problem may still need attention. If nobody fixes it and resets the generator within a given time, the controller will call with a generator running alarm.

“At the main plant site, we are running S7-300 PLCs, which are much more powerful than the S7-200s in the lift stations,” Tankersley says. Additional equipment used in the main facility includes 8” HMI color touch panels, and 19” touch panel HMIs. All remote PLCs automating processes scattered throughout the plant have Siemens Scalance 788-1 PRO radios communicating with the main operations console via wireless Ethernet (WiFi) in the operations building.
There are two PCs in the operations building. One runs Win911 and the other is the WinCC Flexible HMI to supervise all plant processes. That second PC also data logs all plant processes, archiving the data. Because the system is stand alone, all PLCs, HMIs and PCs are synchronized to the exact time through a GPS time server.

Altogether, the system upgrade and wireless network have been successful. The new communications are much more reliable, while the new automation equipment proved simpler to install and easier to maintain than the old equipment. The staff at Persigo can now respond much more quickly to alarms at any of their locations, local or remote, and with more diagnostic information. This cuts down on their travel time for troubleshooting with most initial diagnosis able to be done remotely, eliminating wasted efforts for nuisance alarms. The new Siemens TIA based system provides an architecture that is compatible for easy future expansion, protecting the investment by Persigo and ensuring future sustainability.

Persigo Waste Water Treatment Plant at a Glance
The treatment process at the Persigo Waste Water Treatment Plant includes four steps:
1. Preliminary Treatment
2. Primary Clarifiers and Activated Sludge
3. Final Effluent
4. Sludge Processing
Design capacity for the plant is 12.5 million gallons per day of wastewater generated at homes, schools, churches, and industries, anyplace that is tied into the sewer system. The wastewater travels to the treatment facility through 462 miles of pipelines. Average flow is 8.0 million gallons/day (30,283 m3/day). It takes about 14 hours for the water to go through the entire plant. In the main facility, there are five sewage lift pumps of 100 horsepower with a capacity of 6,950 gallons/min (10 million gallons/day) each.

Read Full Post »

Precision manufacturing is key to producing weather-resistant solar panels. Integrated motion and control systems from Siemens provide the needed accuracy and performance.

A market growing at 45 percent per year looks like a good bet to get into. According to Mike Taylor, director of research and education, Solar Electric Power Association, the solar photovoltaic for electric generation market has been on that slope for the past three years. No wonder automation companies are developing solutions for the solar panel market.
Taylor says that Europe has been the photovoltaic leader, but that the U.S. market is rapidly developing. Putting to rest the myth that solar implementations require never-ending sun, the current market leader is Germany—not exactly a sunny country. Spain follows, with the United States third in market size.
“In the United States, we just broke the 300 megawatt (MW) barrier in photovoltaic electric generation in 2008. By comparison, in 2007 the total was 160 MW and in 2006 it was not quite 100 MW,” says Taylor. “With the entry of utilities and big box commercial stores into the (solar energy use) market, it’s just going to keep growing.”

Siemens has been involved in the solar industry for many years and has honed its automation offering to fit the needs of solar panel manufacturers. Apparently simple from the outside, these panels are multiple layers of materials that must be precisely sized, cut, stacked and sealed. If dirt, dust or contaminants land on any of these layers before the unit is laminated, it more than likely will affect the performance of the solar panel. The assembly process must assure the final product is impervious to weather and other outdoor environmental hazards.

Complex manufacturing
The manufacturing process includes accurately measuring, cutting, transporting, positioning and stacking insulation, protective film and other layers of proprietary films onto expensive and fragile sheets of photovoltaic silicone without contaminating any of the layers. If the silicone layer is dropped, bumped or hit with too much force, the sheets can crack or break into shards. The films come from rolls and are cut to length and registered to position onto the substrate. The end result is a hermetically sealed panel that converts solar energy into electricity for residential or commercial use.
A company located near Chicago specializes in building machinery to manufacture these panels. A Siemens 317T programmable logic controller

“This was a new project for this customer,” says Ken Brey, technical director for DMC Inc., an automation engineering and software services company located in Chicago. “This customer requested Siemens equipment to match the rest of their automation equipment. They know that Siemens equipment would not only give them continuity in the hardware and software, but would also give them the high level of performance and operation they expected. They also know from experience that Siemens service is much easier to get globally than other brands.”

(PLC) controls the process. This PLC employs CPU 317T-2 DP technology which adds powerful motion control functions to the existing range of standard central processing units (CPUs). These motion control functions are integrated in the CPU firmware and are supported by a co-processor for high-speed closed-loop control. This ensures that the performance of the CPU 317 for normal control tasks is maintained and that the motion control functions can be computed at a higher clock rate.

Like making a sandwich

As Brey explains it, the process of making these panels is similar to making a sandwich. Instead of bread, glass panels make up the first layer. Instead of meat or cheese, EVA and insulation material make up the second layer. The solar panels get placed into the sandwich and are put into a laminator to seal everything together.

“Our customer’s machines cut and place the middle layers onto a vacuum conveyor. The machine positions the panel under the leading edge of the EVA and feeds the panel and the EVA together to build our “sandwich.” The laminating process is done with another company’s equipment,” states Brey. “The end customer was building an entire factory from scratch with equipment from different vendors.  Standardizing on Siemens was the quickest and easiest way for the equipment to communicate as well as ensuring the job met the customer’s high expectations,” according to Brey.

A typical motion control application like this assembly process needs to control 2 to 8 axes. The Siemens 317T PLC can control up to 16 axes. In addition to accurate single axis positioning, the technology CPU is particularly suited for complex, synchronized motion sequences, such as coupling with a virtual or real master, geared synchronous motion, electronic cam discs and print mark correction. Print mark registration is particularly important in the solar panel lamination process so film is properly aligned onto the photovoltaic cells. Using Siemens STEP 7 software, only one program is needed for PLC and motion control, making configuration and programming significantly easier and faster.
The processor in the solar panel assembly machine has motion control functions integrated into the CPU that guide four servo axes to position the film. It allows the process to run at high speeds without a separate motion controller. Accurate registration of the film is achieved using the built-in geared synchronous motion and measuring probe functions of the 317T PLC. Siemens solution provider, DMC, Inc., installed the control system at the manufacturer’s facility near Chicago and the machine now resides in the photovoltaic panel plant in Europe.

SEPA’s Taylor has calculated from public announcements of utilities that 2,500 MW of electric generation from photovoltaic panels is projected for manufacture in the next five years. With utilities and large commercial enterprises such as Wal-Mart looking at adding solar generating equipment, the demand for these Siemens controlled assembly machines should be sky high.

For more information on Siemens Solar Industry Solutions, please click here.

Read Full Post »

The latest LOGO! AM2 RTD module makes it simpler for users to add a range of sensors. The new module provides improvements that include support of PT100 and PT1000 sensors while also letting users mix either PT100 temperature sensors or PT1000 Flue gas temperature sensors.
The LOGO! AM2 RTD is fully compatible with the existing products (LOGO! 24, LOGO24o and AM2 PT100) and will replace the predecessor products. The line simplifies setup: when new sensor types are connected, the modules provide automatic identification of the connected sensor when the expansion module starts up.

LOGO! modular is available in various versions for a variety of supply voltages (12 V DC, 24 V DC, 24 V AC, 115/230 V DC, 115/230 V AC), offering distinctive features such as R: Relay output and C: Clock/time switch. Some LOGO! Versions include operator keypad and display panel in one unit, streamlining system design. Up to 36 different functions can be connected at the click of a button or by means of PC software.

For more information on the LOGO! AM2 RTD, please click here.

Read Full Post »

When overloads and short circuits occur, there’s always a tradeoff between the power supply shutting down the entire 24 VDC supply and keeping load circuits that aren’t impacted by the failure continuing to run. The new SITOP PSE200U electronic diagnostic module from Siemens Industry Automation Division constantly monitors up to four 24 VDC load circuits for overloads and short circuits.
“Today’s switched-mode power supplies provide advanced functionality, by protecting 24 VDC loads against any kinds of overload. Unfortunately, it responds by shutting down the power supply to all load circuits even before a fuse or circuit breaker can trip on the circuit that is actually being overloaded. The result: Power is shut down to all 24 VDC load circuits, even those that aren’t affected,” says Kai Bronzel, product marketing manger, Siemens Industry. “The SITOP PSE200U electronic diagnostic module intelligently shuts down the current to the affected load circuit, without impacting the other load circuits that don’t have problems.” Furthermore, the SITOP PSE200U provides for sequential start up, which significantly reduces the total current required from the power supply during the start up phase of high inrush loads. This avoids the shutdown of the power supply when the total inrush current exceeds the current limitation of the power supply.
In order to more closely match the application, there are now two versions of the device with tripping current ranges of 0.5 to 3 and 3 to 10 amperes available. This is an improvement over the previous SITOP select model, which had a wide adjustment tripping current range of 2 to 10 amperes. Used as an additional component in switched mode regulated 24 VDC power supplies, the new SITOP PSE200U electronic diagnostic module detects not only slight overloads but also “creeping” short circuits on high-resistance cables. If an overload or short circuit occurs, the device switches off the current to the faulty path. This prevents the power supply voltage from dipping, thus ensuring that power continues to be supplied to all the other circuits not affected by the fault. Up to four load circuits per module can be protected. Additional electronic diagnostic modules can be connected to the power supply if the application requires the protection of more circuits.
The new SITOP PSE200U electronic diagnostic module has sealable potentiometer covers to protect the set tripping currents against unwanted changes. The three-color LED and signaling contact concept makes it simple to troubleshoot systems and get directly to the faults. Diagnostics data can be transmitted to a higher-level controller for remote diagnostics. The SITOP PSE200U allows fast commissioning due to switch on/off functionality by using individual reset buttons per circuit in front of the unit. In addition a remote reset from a higher level control system is also possible. In contrast to conventional miniature circuit breakers, which do not trip until the current rises to a multiple of the rated current, this new electronic diagnostic module already responds to an overload that is only slightly above the set tripping current. The device also differentiates between a sustained fault and a transitory, operational overcurrent. This enables short circuits to be detected even on high-resistance cables in which, for example, core cross-sections are small or there is a long distance to the load.

Click here for more information on SITOP PSE200U electronic diagnostic module.

Read Full Post »

Engineering students at Virginia Tech’s Robotics and Mechanics Laboratory (RoMeLa) recently unveiled the first complete, untethered, autonomous walking humanoid robot built in the United States. The students used NI LabVIEW and NI Single-Board RIO to create the robot.

The Cognitive Humanoid Autonomous Robot with Artificial Intelligence, better known as CHARLI, was conceived by RoMeLa associate professor and director, Dr. Dennis Hong. The robot was built with only $20,000 in funds and donated equipment from NI and Maxon Precision Motors.
CHARLI’s structure is anatomically based, deploying a system of pulleys, strings, carbon fiber rods and actuators, instead of using traditional rotating joints. Standing at five feet tall, the robot can climb stairs and tread uneven ground, which is more than most robots can handle.
There are two CHARLIs – CHARLI L, as in lightweight, and CHARLI H, as in heavyweight. CHARLI H is still being developed with the hopes that it will one day be able to run, jump, and do just about anything a person can do.

Watch the video to see the robot in action and hear the students talk about CHARLI.

Read Full Post »

Factory Automation with industrial robots for ...Image via Wikipedia

Industrial robots and growth seem made for each other. 

But the growth will be narrowly focused, predicts Erik Nieves. “Robotizing” what’s already automated is one way, says the technology director for supplier Motoman Robotics (www.motoman.com), Waukegan, Ill., a division of Yaskawa America Inc. While that provides flexibility, “effectively, you’re just making an improvement, though probably not in reliability. You’re leveraging the flexibility that robotics affords,” he observes. Whatever changes, though, robotizing of already automated functions “is not going to be the ‘sea change’ in our industry.”

Another trend he sees, unified control, could be part of such change. In this control infrastructure, especially in emerging markets that haven’t existed long enough to assume separate controls is correct, “you make a decision about who’s responsible for everything,” Nieves explains. “Can I program not just the robot with the robot controller but all the peripherals?” If so, there won’t be a programmable logic controller (PLC) involved, he continues. This control unification for robots hasn’t yet hit the factory floor, he says, but it’s coming.

What Nieves considers the real “sea change” is dual-armed robots. These perform tasks with the dexterity previously possible only with humans, he comments. The big leap was not from a single-arm to a dual-arm robot, though, he asserts. “That was natural. The leap was from six to seven axes. Once you have a seven-axis arm and you have all this dexterity, then it is very organic to apply two seven-axis arms to a common torso.” And that, Nieves states, was how the dual-arm robot was born.

The seven-axis arm allows movement without affecting position and orientation, he adds. And such dual-arm robots provide gains in productivity. But these robots, which are in their infancy, will never be dominant, Nieves predicts. “You will always have tasks that six-axis robots are suitable for, such as tried-and-true applications such as spot welding.”

Besides more mechanical functionality, robotics’ growth also focuses more on interconnectivity and traceability, suggests Rush LaSelle, global sales and marketing director for vendor Adept Technology Inc. (www.adept.com), Pleasanton, Calif. One pathway is through information-technology clouds. This technology implies that the computing resources exist somewhere else, “out there,” and, as necessary, that users will connect. “The benefit of processing and its interrelationship with the cloud is in its relative infancy throughout manufacturing and, certainly, within the context of robotics.” But, says LaSelle, “ ‘cloud’ or resource sharing has been used in the automation industry for as long as there have been networks.”

Cloud coverage

Another real-world cloud example he provides is inventory control, in which robots palletize and handle products in warehouses and distribution centers. “Inventory information is largely managed in the cloud and the addition of intelligent robotic automation acts as a vehicle to reconcile what’s ‘expected’ in the digital world—that is, what is captured in databases throughout planning and finance—and what actually ‘exists’ in the physical world—that is, pallets of products in warehouse racks,” LaSelle explains.

Besides clouds, end-users will benefit from being freed of having to program robots. “Our team is evaluating technologies and developing methodologies so that users will be able to teach, not program, robots,” LaSelle states. Options he mentions include gesture-based teaching and other human-machine interfaces beyond the traditional interface on a teach pendant, programmable logic controller or personal computer.

Sensory inputs comprise another growth area of value-add focus. “Portion control and product grading for the handling of primary foods, especially those in the protein category (e.g. meat, fish and poultry), are active areas of development,” LaSelle notes. What could the net benefit be? Robots will not only locate, pick and place products into packaging, he says, but will determine size of the products. That’s so the robot can place the correct weight of products into a package and conduct real-time quality control.

C. Kenna Amos is an Automation World Contributing Editor.

Motoman Robotics

Adept Technology Inc.

Enhanced by Zemanta

Read Full Post »

The Wi-Fi certified logo found on many Wi-Fi e...Image via Wikipedia

Wireless Fidelity, or Wi-Fi, technology is proliferating in the factory, but it’s not the best answer for everything. Will Wi-Fi take over the industrial world? How about wireless Ethernet? The second question is easy to answer: there is no wireless Ethernet. Ethernet, IEEE 802.3, runs on wires. Yes, you will see the term. The current generations of Wi-Fi, or Wireless Fidelity—IEEE 802.11a, 802.11b, 802.11g and 802.11-2007 (which rolls up -a, -b and -g with the lesser-known -h, -i, and -j)—define wireless local area networks (LANs) that have become so inextricably linked with Ethernet that many call Wi-Fi, “wireless Ethernet.”

Wi-Fi is taking over a range of factory applications. Part of the reason might be called peer pressure: outside of industrial settings, there is a huge installed base of IEEE 802.11 LANs—based on the Institute of Electrical and Electronics Engineers standard—at work literally everywhere. A second driver is the direct link to Ethernet, both for environments that use industrial Ethernet and for communication with Ethernet-based enterprise IT systems.

The reasons for the widespread usage of Wi-Fi are many. First, the silicon investment is minimal on the commercial or consumer side. Chips and circuits are commodity items at commodity prices, driving equipment prices down to less than a single meal at a moderate restaurant.

Second, people like Wi-Fi and demand it in their solid-state goodies. Name any reason from sternly practical to frivolously air-headed and someone around you will be trying to connect using 802.11(x). The result is that every new laptop computer—in fact, any new device that will reach into e-mail or Web addresses—includes wireless, and the lines between cell phones, personal data assistants (PDAs), MP3 music players and wireless computers are increasingly blurred.

The upshot is that, while Wi-Fi was once a struggling new technology, it is now literally easier to log on to a wireless LAN anywhere than to avoid logging onto a LAN. In fact, the 802.11 airwaves are now so crowded that many of us (especially those on the road) spend appreciable time figuring out just which LAN we are using at a given locale. It can take many minutes to ensure that you are on the LAN you want amid all the LANs around you.

Traffic flowing on 802.11 highways is likely to be in the air around you, especially if your facility is anywhere near office buildings or family residences, or if your corporate IT embraces wireless connectivity (most do).

So, Wi-Fi is beckoning to production. Broadly speaking, there are only three responses to its siren call. The first is to simply hold off while the current recession blows chill winds through every kind of technology, especially those committed to silicon chips. It is unclear who might benefit from this approach (if anybody). For a while, there will be fewer changes to keep up with. In the vendor community, the survivors will be smaller, leaner and hungrier than last year’s (or last decade’s) boom-time participants. Unfortunately for manufacturing, at the same time, the availability of resources for custom installation will be greatly diminished—and manufacturing absolutely depends on customization.

The second is to hold off until the next generation of Wi-Fi specification, IEEE 802.11n, becomes mainstream. A new generation, 801.11n may reach finalization in November of this year. Its promise—less interference, more data throughput, possibly enhanced security—provides a rosy glow for the future, a glow that will almost certainly invite industrial needs into its warmth. More on that later.

The third is to evaluate current Wi-Fi in relation to factory needs. Then if the technology is appealing, the next steps are familiar from any network implementation: study, strategize and install.

But these are not the only responses. “Wi-Fi and Ethernet are solidly entrenched technologies, but there are better choices for applications such as sensor networks,” says Cliff Whitehead, manager of strategic applications at Rockwell Automation Inc., the Milwaukee-based automation vendor. Whitehead is co-chair of the factory automation study group of the International Society for Automation’s ISA100 standards committee, which is developing an industrial wireless standard. “Remember, radios were used in manufacturing long before computers or 802.11, or any comprehensive set of standards, for that matter. The result is that there are many point solutions involving licensed and unlicensed radio bands, cellular or any number of media for sending this or that kind of data without wires. They all work, and for some needs, many of them work better than Wi-Fi.”

If Wi-Fi is having trouble reaching into control networks, one reason is performance. Whitehead points out that performance on existing 802.11a/b/g technology is not as fast as wired. “For periodic monitoring, say every second or so, Wi-Fi works fine,” Whitehead says, “but for high-speed motion control in microsecond time frames, wireless is not there in a/b/g, and it pushes the envelope in [802.11]n. On the other hand, for peer-to-peer data sharing, or for mobile workers with a laptop doing program adjustments and troubleshooting, Wi-Fi is an excellent alternative.”

Hesh Kagan, managing consultant, enterprise architecture and integration, for automation supplier Invensys Process Systems, in Foxboro, Mass., and president of the Wireless Industrial Networking Alliance (www.wina.org), agrees that Wi-Fi is far from the answer to everything. “There are two major divisions in the industrial wireless world,” he says. “The first applies to workstations or devices in enterprise-wide or secondary implementations using 802.11 Wi-Fi. These are not directly involved in control. The second revolves around field sensors as part of operational control. In this arena, Wi-Fi unfortunately has a big footprint and requires huge amounts of power compared to the far less power-hungry equipment designed to meet ISA100 and 802.15.4. Battery life is extremely important in the field sensor world, and Wi-Fi would suck batteries dry quickly.”

If sensor networks and control applications are not ideal for Wi-Fi, Kagan suggests several layers of applications that are well-suited, ranging from least complex to most. In general, the applications are adjuncts to (rather than direct participants in) process instrumentation or machine control—that is, they provide overview and management functions rather than operational control.

The first of this kind of adjunct or helper application is wireless video: “Remote visualization provides the easiest application,” Kagan says. A read-only application, Wi-Fi-based video takes advantage of a broad array of low-cost products. “There are highly capable cameras for low light and external applications,” he explains. “They are easily set up and provide a dead simple way to gather images of whatever needs to be watched, whether that be perimeters for intruders, or tanks for leaks, or any number of safety-related needs. You can even focus on dials and indicators if you want to lighten the load on a roving clipboard-carrier.”

Neil Peterson, services marketing manager for the wireless plant network, at vendor Emerson Process Management, Austin, Texas, agrees: “Video monitoring offers cost-effective visualization of things like emissions, perimeter control and safety needs. Because Wi-Fi shares resources with a variety of network configurations, it can share much of existing infrastructures. That makes installation relatively painless.”

Similar applications include Wi-Fi-enabled motion detection, ambient heat or carbon monoxide sensors and similar devices with radios. All of these devices are well established as commercial security and safety units.

Slightly more complex is wireless enablement of mobile operators, who can benefit from two-way hookups. Here, specific operational information is channeled to end-users, giving them access to data or information that in a non-wireless world would be locked away in printed manuals or fixed-station computer terminals. At the same time, operational or maintenance data can travel from the mobile resource back to the control rooms, enterprise systems or remote resources via the Web.

“There is value in untethering people from the control room,” Kagan says. “Inter-process measurements, operator access to set points or remotely acknowledging alarms—any number of mobile applications—are ideal for Wi-Fi-enabled handhelds or computers. A degree of security is involved, but it’s easy enough to allow or disallow specific changes to specific operators.”

“[Wi-Fi] access points are a means to an end, and the end in this case is mobility,” says Emerson’s Peterson. “Wi-Fi offers a cost-effective way for an operator to run things from a hand-held while performing manual steps in the field. In small plants especially, Wi-Fi allows people to break the chains to their desks.”

“The next layer up, control over mobile assets, is gaining a lot of interest,” Kagan says. “RTLS (real-time location services) benefits from Wi-Fi-enabled tags on equipment or badges on people. Setting up mobile asset tracking this way has no impact on the processes or controls, and it allows you to add important capabilities with very little outlay.”

“With wireless RFID [radio-frequency identification] tags, you can know exactly where your personnel are,” Peterson points out. “This can be valuable for teams entering hazardous areas—you can keep an eye on exposure time. More importantly, in a safety-related mustering, RFID tags let you know exactly who has exited dangerous areas and who has not.”

Note that it is easy enough to avoid intrusive tracking, simply by deactivating the system except for those moments when safety demands require a clear, accurate picture of where people are.

A final touch is the ability to use Wi-Fi to send relevant information to maintenance teams. “You can codify a repair procedure into step-by-step tasks,” Kagan says. “Then, wherever they are, they can positively identify a piece of equipment, then download the exact procedures for a given fix. And you can do more than that. You can collect machine data and repair information and send it back up to the server, again, from anywhere within reach of [a Wi-Fi] access point. The result is a maintenance database that reflects reality.”

As part of this vision, last August, the Wonderware business unit of Invensys acquired the Houston-based SAT Corp., with its IntelaTrac Enterprise Suite set of mobile offerings. They include configurable software and mobile hardware for workflow, procedural and general task management. Originally focused on maintenance, IntelaTrac is expanding into broad-based production and compliance applications.

Looking for more complexity? Thanks to the tight integration of Wi-Fi and Ethernet, the most widespread LAN technology, there should be no limit on application areas for which the latter offers advantages. Ironically, Ethernet’s original inspiration was radio broadcasting, enhanced by the capability to detect data collisions better in wired connections than in radio frequency connections.

Once data is gated over to Ethernet, the first gain is the ability to communicate with a huge array of devices running a broad range of applications. Enterprise information technology (IT), in particular, depends on Ethernet, so a Wi-Fi bridge to Ethernet makes many an IT practitioner feel more comfortable around production data. The second gain is direct connectivity to the growing application base running on Ethernet’s rough-and-tumble sibling, industrial Ethernet (IE). IE as a link-layer protocol offers increased bus speed compared to serial buses, as well as access to relatively low-cost, standard devices. (For more detail, type “industrial Ethernet” into the search box on AW’s Web site, www.automationworld.com. You’ll find a goodly store of features, white papers and product information on the technology.)

As with much of wireless technology, factory Wi-Fi applications are still in their infancy. Peterson points out that ramping up offers the greatest flexibility. “It’s easy enough to begin with a scattering of access points,” he says. “You might start with a single backhaul network or a wireless field network, maybe even funneled through a single access point to the control room. In this way, you gain a hot spot for mobile worker coverage or for any number of uses. Next year, another access point can be added for relatively low cost, then another. Each addition broadens Wi-Fi accessibility without drawing too much budget.”

“One emerging growth area is a potential for productivity gain through the reining-in of your engineering drive for pinpoint precision,” Kagan says. “Instead of a $1,500 temperature monitor that resolves to one tenth of a degree over its entire range, how about a relatively coarse wireless unit on a motor just to see if it is overheating? Surrounding a potential problem area with many casual monitoring sensors can prove more practical than employing one or two high-precision devices. Clearly, this kind of thing is reserved to situations where a few seconds lag time or a few degrees off true reading won’t make a difference, but there are a lot of these out there.”

Progress on IEEE 802.11n standards is adding a new layer of potential applications. Already there are “Draft n” devices in the marketplace. The developing standard promises to accelerate the rate at which Wi-Fi gains adherents. For one thing, 802.11n offers significantly higher throughput with significantly reduced latency compared to earlier 802.11 specifications.

“The biggest gain is increased reliability in transferring data quickly and completely,” says Whitehead. He cites a number of technical elements in 802.11n that enhance transmission, chief among them MIMO (multiple in, multiple out) technology that takes advantage of multipath signals and multiple antennas. The approach enables the transmission of significantly more information than is possible with single antennas.

“Conceptually, you’re transmitting simultaneously on multiple radios,” he explains. “In our testing with draft n Cisco equipment, we’re seeing 2,400 to 7,500 packets per second. That’s approaching rates consistent with discrete I/O (input/output) control. ‘Real time’ has many meanings in manufacturing, but this is near real-time for many uses.”

Additionally, while previous 802.11 standards focused on 2.4 gigahertz (GHz) frequency bands, 802.11n can use 5.8 GHz frequencies. “There’s less congestion in 5.8,” Whitehead says, “so there are fewer latency or error problems traceable to the coexistence of competing transmissions.”

Finally, 802.11n incorporates a number of other new approaches. Channel bonding (transmitting information on two non-overlapping channels) further speeds up data rates. Data encoding and aggregation algorithms help decrease overhead while increasing signal clarity and speeds. Through it all, the standards process has emphasized backward compatibility with earlier 802.11 specifications, though the overall throughput is necessarily reduced when in compatibility modes working with older Wi-Fi devices.

Meanwhile, to step from the future back into the present, any enablement or enhancement to factory life through Wi-Fi, regardless of how cutting-edge or exploratory it might be, depends on standard processes for technology planning and implementation.

“It boils down to the same three things, whether you’re dealing with Wi-Fi, wireless Hart, wired networks or a person with a clipboard,” says Whitehead. “The first is performance, the second is reliability and the third is security. Wireless doesn’t have the same inherent reliability as a bundle of cables in many an application. Plus, if you decide that wireless is the solution, you need to remember that performance and reliability are affected environmental factors. You need to know what other wireless co-exists with your installation, since existing radio traffic has the potential for radio frequency overlap. And you have to remember that if Wi-Fi drops packets, it will cycle through retries. While 30 seconds one way or the other won’t bother you if you’re downloading an MP3 audio file at home, in a control situation with time-outs built into the protocol, you risk nuisance trips or downright outages.”

Standard procedures are emphasized by Peterson as well. “Every solution involves software, hardware, services and customization, and each one has to be scrutinized,” he says. “What do you want to do? What will you want to do in the future? You need to involve the stakeholders first to see what needs to be done. Then you evaluate the site, and for Wi-Fi, you’ll want to do a radio frequency site survey to see what’s in the air, so to speak. Then it’s just a matter of installing the wireless, the applications, and checking it all out. A finished installation will include planning for after-install services—training, maintenance, ongoing evaluation. You can’t just say, ‘Here’s a box, have fun.’ ”

There are a range of questions around architecture, focusing first on the exact location of the radio, then on the transfer of data to whatever end-point is chosen.

“You’re looking at a virtually infinite number of ways to put systems together,” says Whitehead. “As the application space settles down and people share the basic concepts, these will no doubt resolve to a few different primary strategies. Right now, however, I doubt that anyone would feel comfortable starting from scratch, working on their own. Everyone in the space—suppliers, consultants, device makers, end-users, everyone—has the responsibility to make sure they are deploying things they understand. Standards are increasingly fleshed out with reference architectures that provide suggestions for deployment. Plus, as standards proliferate, more and more people will be comfortable with the technologies involved. Once that happens, you’ll find more and more people able to see both the capabilities and the shortfalls of a given approach.”

Wireless Industrial Networking Alliancewww.wina.org

Enhanced by Zemanta

Read Full Post »