Order ID | 53563633773 |
Type | Essay |
Writer Level | Masters |
Style | APA |
Sources/References | 4 |
Perfect Number of Pages to Order | 5-10 Pages |
Software Integration and Advanced Control Essay
ROCKWELL AUTOMATION’S CHIEF TECHNOLOGY OFFICER, SUJEET CHAND, PONDERS WITH MANUFACTURING ENGINEER OVER WIRELESS, RFID, SOFTWARE INTEGRATION AND ADVANCED CONTROL, JUST A FEW OF THE ELEMENTS THAT
WILL SHAPE THE FUTURE OF FACTORY AUTOMATION. By Sean Davies
Unlocking the future of
automation
T here can be few companies better placed than Rockwell Automation to gaze into the crystal ball and predict the technologies that will alter the automation landscape over the coming years. With an annual turnover in excess of $5bn, and upwards of 21,000 employees, the
Milwaukee, Wisconsin-based automation company – much like its rivals Siemens, ABB and Invensys – is attempting to shape the future of automation and control. It is watching the rise of wireless technology with interest, tracking the rising profile of RFID and following the evolution of the software generation. It is plotting all these trends against an ever changing manufacturing landscape that is demanding greater flexibility, an increasingly integrated IT infrastructure, and is in the throes of seemingly endless supply chain integration.
As chief technology officer, Sujeet Chand is the man charged with aligning product and technology strategies with market needs. He confirms that Rockwell has been collating technology trends for many years now; tracking roughly 80 technologies on an annual basis. “What we’ve done is we’ve taken these 75 or 80 technologies and created five clusters, five major technology clusters, that impact industrial automation,” Chand says. “However, technology trends are not of any relevance unless they solve problems in manufacturing – problems that our customers perceive in what they do.”
The five technology clusters Rockwell monitors are control and diagnostics, communications, electronics, materials and software. When considering the problems facing manufacturing, Chand ponders before mentioning the integration of factories within business systems such as SAP, a move driven by supply chain integration and
E Manufacturing Engineer | February/March 2006
flexible manufacturing or customer demand. Within the factory itself, issues such as regulatory
compliance drivers, safety and security, conformity and certainly cost and energy efficiency are top of his list.
“We hear about change from mass production to mass customisation – everything from food to pharmaceuticals to automobiles,” he says. “We’re now talking about customised manufacturing – Mills experimented with creating customised cereal on a website called mycereal.com, where you could go in and say, ‘I want one third bran flakes, one third cornflakes, and one third raisins’, and mix that for you and deliver it to you. That’s an example of customised manufacturing.”
So how is the control and automation sector reacting to the shifting backcloth? The answer is by adopting, or, more accurately, incorporating emerging technologies – a good case in point is industrial Ethernet. Ethernet is the fastest growing network in industrial automation and, according to Chand, it is going to become the dominant network in the next few years. In addition to this we are also seeing the rapid adoption of the world wide web software standards such as SML being used on the factory floor.
INTRODUCING ADVANCED CONTROL To marry these advances in technology with the demands of manufacturing, Chand explains that what is required is advanced control. “Today, PID (Proportional Integral Derivative) is ubiquitous in industrial automation; regardless of what problem it is you use PID control,” Chand explains. “Increasingly, our customers are looking for more advanced control. Many of our customers are looking to apply the major advances that we’ve made ➔
IEE Manufacturing Engineer | February/March 2006 13
flexible manufacturing or customer demand. Within the factory itself, issues such as regulatory
compliance drivers, safety and security, conformity and certainly cost and energy efficiency are top of his list.
“We hear about change from mass production to mass customisation – everything from food to pharmaceuticals to automobiles,” he says. “We’re now talking about customised manufacturing – Mills experimented with creating customised cereal on a website called mycereal.com, where you could go in and say, ‘I want one third bran flakes, one third cornflakes, and one third raisins’, and mix that for you and deliver it to you. That’s an example of customised manufacturing.”
So how is the control and automation sector reacting to the shifting backcloth? The answer is by adopting, or, more accurately, incorporating emerging technologies – a good case in point is industrial Ethernet. Ethernet is the fastest growing network in industrial automation and, according to Chand, it is going to become the dominant network in the next few years. In addition to this we are also seeing the rapid adoption of the world wide web software standards such as SML being used on the factory floor.
INTRODUCING ADVANCED CONTROL To marry these advances in technology with the demands of manufacturing, Chand explains that what is required is advanced control. “Today, PID (Proportional Integral Derivative) is ubiquitous in industrial automation; regardless of what problem it is you use PID control,” Chand explains. “Increasingly, our customers are looking for more advanced control. Many of our customers are looking to apply the major advances that we’ve made
12 IEE Manufacturing Engineer | February/March 2006
T here can be few companies better placed than Rockwell Automation to gaze into the crystal ball and predict the technologies that will alter the automation landscape over the coming years. With an annual turnover in excess of $5bn, and upwards of 21,000 employees, the
Milwaukee, Wisconsin-based automation company – much like its rivals Siemens, ABB and Invensys – is attempting to shape the future of automation and control. It is watching the rise of wireless technology with interest, tracking the rising profile of RFID and following the evolution of the software generation. It is plotting all these trends against an ever changing manufacturing landscape that is demanding greater flexibility, an increasingly integrated IT infrastructure, and is in the throes of seemingly endless supply chain integration.
As chief technology officer, Sujeet Chand is the man charged with aligning product and technology strategies with market needs. He confirms that Rockwell has been collating technology trends for many years now; tracking roughly 80 technologies on an annual basis. “What we’ve done is we’ve taken these 75 or 80 technologies and created five clusters, five major technology clusters, that impact industrial automation,” Chand says. “However, technology trends are not of any relevance unless they solve problems in manufacturing – problems that our customers perceive in what they do.”
The five technology clusters Rockwell monitors are control and diagnostics, communications, electronics, materials and software. When considering the problems facing manufacturing, Chand ponders before mentioning the integration of factories within business systems such as SAP, a move driven by supply chain integration and
➔
ROCKWELL AUTOMATION’S CHIEF TECHNOLOGY OFFICER, SUJEET CHAND, PONDERS WITH MANUFACTURING ENGINEER OVER WIRELESS, RFID, SOFTWARE INTEGRATION AND ADVANCED CONTROL, JUST A FEW OF THE ELEMENTS THAT
WILL SHAPE THE FUTURE OF FACTORY AUTOMATION. By Sean Davies
Unlocking the future of
automation
there are only two key differences between bar codes and RFID,” Chand says. “The first is that a bar code requires line of sight – RFID does not. A second key difference is that the bar code technology is designed for tracking batches. For example, if I pick up that bottle of water, there’s a bar code on it – if I scan it, it tells me it’s a bottle of water made by that manufacturer. It doesn’t tell me anything about this specific bottle of water, with RFID, with 96 bytes you can do a lot more.
“When they discovered that they could marry identification technology to the Internet, they thought it was revolutionary. It doesn’t matter whether it’s RFID or any other identification technology. So what’s so great about marrying identification with the Internet? Every time you see that identification, you can put information on a website. That’s why we think of RFID as a train, every time it makes a stop; a new part gets added on to the train. So this train of information keeps growing – you can add different bits of information associated with a single tag. That’s the difference, that’s the key value that RFID brings.”
However, even with those advantages just using RFID, it’s not sufficient. If RFID is used to change processes, then you can get tremendous value from it, although, as Chand points out, there are still limitations, such as reading the tags in difficult conditions – on metal objects or near liquids. On top of that there are still some standards issues that need to be resolved, although the IEEE is working on these. But, in reality, the biggest obstacle is still the high cost compared to prospective return on investment. “When we started thinking about return on investment, one of our customers said, ‘it’s a cost of $200m a year with zero return’,” Chand explains. “Is Wal-Mart going to pay more for it? No. You’ve got to find efficiencies in your process to get that value, and based on our analysis to get the value from RFID, you’ve got to move the tags back into the factory. You’ve got to find the tag earlier – the earlier you find the tag, the more return on investment you can justify.
“For example, if you apply the tag in the factory itself, not only do you associate all of your process data with that single identifier as opposed to creating a separate identifier and creating two separate databases for the same product, you can track shrinkage, you can dump products based on RFID tag information. So we see RFID moving back into the factory.”
DRIVEN BY SOFTWARE It has been recognised for some while that within factories the traditional three tier IT structure of enterprise, manufacturing and control has been compacting, with the two lower tiers merging. Again supply chain integration is a key driver along with the two-way flow of design information. “You’ve got design, you’ve got product design where designers are designing the product and you’ve got support – it used to be a one
way track. You design the product, you push it out of the factory, warranty is somebody else’s problem. Now when you have a warranty issue, you want to look back into your factory, your manufacturing process, find out what caused the problem, you do your Six Sigma analysis, you feed that information back to design. So then design and support are coming together and guess where they’re leading in manufacturing.
“So you’ve now got the factory and the enterprise coming together, and in the middle is manufacturing. So this implies software, this shows you how important software’s going to be in the future of industrial automation. Today you have enterprise level systems at the high level and in the middle you’ve got the manufacturing execution systems (MES). What we are trying to do here is to integrate multiple control disciplines down at the factory level, and at the moment at the MES layer you’ve got multiple production disciplines that are not integrated. Our strategy, our long-term goal is to really bring these two together so now you have factory data integrated with MES, using open standards.”
There are obvious barriers to software integration in the MES layer, probably hardest to overcome is the fact that it is rich in legacy and also rich in multi vendor applications. Undoubtedly there will be surprises on the horizon, some driven by the vast amount of knowledge that is being collected and stored by all sorts of control devices on the factory floor, others by improved products, but the greatest achievement will be true open standards of devices, software codes and communi- cation protocols. ■
➔
If you only focus on control as a thin slice, you’re only going to get so much out of it Sujeet Chand, chief technology officer, Rockwell Automation
‘‘ ’’
012-015_ME_FebMar05_EY.qxd 13/2/06 15:54 Page 14
in control to their processes, and what they’re seeking is more efficiency, they’re seeking more productivity, they’re trying to get more out of what they do.
“When I gave a talk to International Federation for Automatic Control, my key message to control theorists was to look beyond control. Look at combining diagnostics, prognostics and control – that’s the future, because if you only focus on control as a thin slice, you’re only going to get so much out of it. If you develop complex models of the physical process, you can also do diagnostics. The same model that you use for control can also be used for diagnosis, when you combine the two, the power is quite significant.”
Rockwell customers like UPS have fluctuating demand, in their case brought on by the holiday season. In September of every year UPS has to double the available capacity and then in February, they tear it down because demand declines. “They asked us, ‘can you come up with a process where if we add additional sorting lines to conveyor lines, the system would self configure, without having to write a lot of custom software?’,” Chand says. “They want to move from fixed automation to work a more flexible automation line that is easy to expand and more tolerant. One technology that holds a lot of promise here is the use of autonomous agents. An autonomous agent – there’s no magic to it, essentially it’s a software agent – like an object and you encapsulate some intelligence in the software artefact that allows it to negotiate with other agents.”
Chand is clear on the point that autonomous agents are no panacea. The problems that it’s most suited to are problems where there is redundancy. If there are multiple ways of getting from point A to point B, agents can negotiate and optimise the best way to achieve this.
In one particular example that Chand uses to illustrate the technique, the US Navy came to Rockwell Automation with a problem on the ship – the Navy looks at a ship as a big floating factory. When they launched missiles from
IEE Manufacturing Engineer | February/March 2006
these cruisers, they have to cool the missile launchers, which they achieved by running chilled water through pipes. “There are so many pipes on the ship, there’s a lot of redundancy of getting the water from one place to the other place,” Chand explains. “So there’s added complexity in the system automatically and that was the problem. We were able to solve this problem using autonomous agents technology. We developed a prototype, it was actually tested on a land based simulator that the Navy developed and now we’re working on a project with the Navy and John Hopkins University to validate this technology further and apply it to the Navy ships.” Rockwell is now looking at where this technology would be useful in manufacturing.
IT’S GOOD TO TALK Communications are another area that is shaping the future of automation, mostly in the areas of wireless and RFID. The potential of wireless sensors, actuators and networks have long been recognised but their deployment has been hampered by several problems, the most notable being battery life. With factories containing literally thousands of sensor changing batteries each year or so is a major hindrance. “We need to have a five or six year battery life,” Chand explains. “We need to work with energy harvesting technology because there’s a lot of ability to pick up energy from the movement, then we can use that to power these wireless sensors. That’s not to say wireless is not showing up in factories. We see wireless showing up in several areas like automated guided vehicles (AGVs), GM has a great example of an application of wireless for AGVs.
“It is also great for HMIs, for human machine interface for certain types of applications. For monitoring, if you want to add a sensor to a pump then wireless is perfect. It’s very difficult to add a sensor and wire it into existing communication networks.”
Chand predicts that wireless will flourish for certain applications but widespread deployment of wireless in industrial automation would be negated by the need for mobile radios and power limitations, as well as customer concerns. “When we ask our customers ‘what are your fears about wireless?’ they say ‘interference, coverage, environmental compatibility and security’,” Chand says.
The potential for interference in factories is boundless – everything from fusion lighting that interferes with the radio signal to microwaves and existing wireless infrastructure like telephones. With so much metal in factories, achieving proper coverage is another concern and as Chand notes, security comes up over and over again.
If wireless has a rocky path to tread, the road to RFID adoption is even more treacherous. To begin to understand the complexities involved Rockwell undertook a pilot at its distribution centre where they tagged printed circuit boards. “What we discovered is that
IEE Manufacturing Engineer | February/March 2006 15
Interview
14 IEE Manufacturing Engineer | February/March 2006
in control to their processes, and what they’re seeking is more efficiency, they’re seeking more productivity, they’re trying to get more out of what they do.
“When I gave a talk to International Federation for Automatic Control, my key message to control theorists was to look beyond control. Look at combining diagnostics, prognostics and control – that’s the future, because if you only focus on control as a thin slice, you’re only going to get so much out of it. If you develop complex models of the physical process, you can also do diagnostics. The same model that you use for control can also be used for diagnosis, when you combine the two, the power is quite significant.”
Rockwell customers like UPS have fluctuating demand, in their case brought on by the holiday season. In September of every year UPS has to double the available capacity and then in February, they tear it down because demand declines. “They asked us, ‘can you come up with a process where if we add additional sorting lines to conveyor lines, the system would self configure, without having to write a lot of custom software?’,” Chand says. “They want to move from fixed automation to work a more flexible automation line that is easy to expand and more tolerant. One technology that holds a lot of promise here is the use of autonomous agents. An autonomous agent – there’s no magic to it, essentially it’s a software agent – like an object and you encapsulate some intelligence in the software artefact that allows it to negotiate with other agents.”
Chand is clear on the point that autonomous agents are no panacea. The problems that it’s most suited to are problems where there is redundancy. If there are multiple ways of getting from point A to point B, agents can negotiate and optimise the best way to achieve this.
In one particular example that Chand uses to illustrate the technique, the US Navy came to Rockwell Automation with a problem on the ship – the Navy looks at a ship as a big floating factory. When they launched missiles from
these cruisers, they have to cool the missile launchers, which they achieved by running chilled water through pipes. “There are so many pipes on the ship, there’s a lot of redundancy of getting the water from one place to the other place,” Chand explains. “So there’s added complexity in the system automatically and that was the problem. We were able to solve this problem using autonomous agents technology. We developed a prototype, it was actually tested on a land based simulator that the Navy developed and now we’re working on a project with the Navy and John Hopkins University to validate this technology further and apply it to the Navy ships.” Rockwell is now looking at where this technology would be useful in manufacturing.
IT’S GOOD TO TALK Communications are another area that is shaping the future of automation, mostly in the areas of wireless and RFID. The potential of wireless sensors, actuators and networks have long been recognised but their deployment has been hampered by several problems, the most notable being battery life. With factories containing literally thousands of sensor changing batteries each year or so is a major hindrance. “We need to have a five or six year battery life,” Chand explains. “We need to work with energy harvesting technology because there’s a lot of ability to pick up energy from the movement, then we can use that to power these wireless sensors. That’s not to say wireless is not showing up in factories. We see wireless showing up in several areas like automated guided vehicles (AGVs), GM has a great example of an application of wireless for AGVs.
“It is also great for HMIs, for human machine interface for certain types of applications. For monitoring, if you want to add a sensor to a pump then wireless is perfect. It’s very difficult to add a sensor and wire it into existing communication networks.”
Chand predicts that wireless will flourish for certain applications but widespread deployment of wireless in industrial automation would be negated by the need for mobile radios and power limitations, as well as customer concerns. “When we ask our customers ‘what are your fears about wireless?’ they say ‘interference, coverage, environmental compatibility and security’,” Chand says.
The potential for interference in factories is boundless – everything from fusion lighting that interferes with the radio signal to microwaves and existing wireless infrastructure like telephones. With so much metal in factories, achieving proper coverage is another concern and as Chand notes, security comes up over and over again.
If wireless has a rocky path to tread, the road to RFID adoption is even more treacherous. To begin to understand the complexities involved Rockwell undertook a pilot at its distribution centre where they tagged printed circuit boards. “What we discovered is that
012-015_ME_FebMar05_EY.qxd 13/2/06 15:54 Page 14
there are only two key differences between bar codes and RFID,” Chand says. “The first is that a bar code requires line of sight – RFID does not. A second key difference is that the bar code technology is designed for tracking batches. For example, if I pick up that bottle of water, there’s a bar code on it – if I scan it, it tells me it’s a bottle of water made by that manufacturer. It doesn’t tell me anything about this specific bottle of water, with RFID, with 96 bytes you can do a lot more.
“When they discovered that they could marry identification technology to the Internet, they thought it was revolutionary. It doesn’t matter whether it’s RFID or any other identification technology. So what’s so great about marrying identification with the Internet? Every time you see that identification, you can put information on a website. That’s why we think of RFID as a train, every time it makes a stop; a new part gets added on to the train. So this train of information keeps growing – you can add different bits of information associated with a single tag. That’s the difference, that’s the key value that RFID brings.”
However, even with those advantages just using RFID, it’s not sufficient. If RFID is used to change processes, then you can get tremendous value from it, although, as Chand points out, there are still limitations, such as reading the tags in difficult conditions – on metal objects or near liquids. On top of that there are still some standards issues that need to be resolved, although the IEEE is working on these. But, in reality, the biggest obstacle is still the high cost compared to prospective return on investment. “When we started thinking about return on investment, one of our customers said, ‘it’s a cost of $200m a year with zero return’,” Chand explains. “Is Wal-Mart going to pay more for it? No. You’ve got to find efficiencies in your process to get that value, and based on our analysis to get the value from RFID, you’ve got to move the tags back into the factory. You’ve got to find the tag earlier – the earlier you find the tag, the more return on investment you can justify.
“For example, if you apply the tag in the factory itself, not only do you associate all of your process data with that single identifier as opposed to creating a separate identifier and creating two separate databases for the same product, you can track shrinkage, you can dump products based on RFID tag information. So we see RFID moving back into the factory.”
DRIVEN BY SOFTWARE It has been recognised for some while that within factories the traditional three tier IT structure of enterprise, manufacturing and control has been compacting, with the two lower tiers merging. Again supply chain integration is a key driver along with the two-way flow of design information. “You’ve got design, you’ve got product design where designers are designing the product and you’ve got support – it used to be a one
way track. You design the product, you push it out of the factory, warranty is somebody else’s problem. Now when you have a warranty issue, you want to look back into your factory, your manufacturing process, find out what caused the problem, you do your Six Sigma analysis, you feed that information back to design. So then design and support are coming together and guess where they’re leading in manufacturing.
“So you’ve now got the factory and the enterprise coming together, and in the middle is manufacturing. So this implies software, this shows you how important software’s going to be in the future of industrial automation. Today you have enterprise level systems at the high level and in the middle you’ve got the manufacturing execution systems (MES). What we are trying to do here is to integrate multiple control disciplines down at the factory level, and at the moment at the MES layer you’ve got multiple production disciplines that are not integrated. Our strategy, our long-term goal is to really bring these two together so now you have factory data integrated with MES, using open standards.”
There are obvious barriers to software integration in the MES layer, probably hardest to overcome is the fact that it is rich in legacy and also rich in multi vendor applications. Undoubtedly there will be surprises on the horizon, some driven by the vast amount of knowledge that is being collected and stored by all sorts of control devices on the factory floor, others by improved products, but the greatest achievement will be true open standards of devices, software codes and communi- cation protocols. ■
➔
If you only focus on control as a thin slice, you’re only going to get so much out of it Sujeet Chand, chief technology officer, Rockwell Automation
‘‘ ’’
IEE Manufacturing Engineer | February/March 2006 15
|
||||||||||||||||||||||||||||||||
GET THIS PROJECT NOW BY CLICKING ON THIS LINK TO PLACE THE ORDERCLICK ON THE LINK HERE: https://www.perfectacademic.com/orders/ordernowAlso, you can place the order at www.collegepaper.us/orders/ordernow / www.phdwriters.us/orders/ordernow |
||||||||||||||||||||||||||||||||
|