Dr. Peter G. Martin of Invensys Operations Management, Mr. Dennis Brandl of BRandL Consulting. Mr. Ian Verhappen of Industrial Automation Networks and Dr. John MacGregor of ProSensus Inc. continue their conversation with Walt Boyes, Control's editor in chief.
"I believe," Dennis Brandl says, "that the next revolution in manufacturing is just starting. There are several available or emerging technologies that will make major changes in how we view and control processes and how we manufacture products. The first major technological discontinuity is wireless and mobile. Not only will information be available from any place at any time, the volume of information about our processes will grow as we incorporate wireless sensors into more of the process. The more we see of the processes, the better we can control them. The smart phones of today will become the operators' assistants of tomorrow, looking and listening to the production process to make predictive instead of reactive control moves."
He continues, "The ability to put super-computing resources into local devices means that they will become smarter, will 'learn' and be able to control better than traditional PID control. Control strategies that are too complex to model from first principles will be solved by learned heuristics, with minimal human intervention. These will give us the means to continue the 2% to 3% per year productivity growth that is the hallmark of the manufacturing industries."
He concludes, "Another major discontinuity is the rapid development of 'additive' manufacturing. This term includes 3D printers and other forms of next-generation manufacturing machines. These devices use printer technology and architectures to build devices and machines, one millimeter—or smaller—layer at a time, reducing waste and intermediate inventories. This will provide major savings to manufacturing companies and end users. When this new technology is combined with 3D modeling software and powerful desktop computers, there will be an explosion of new products and ideas that can be built. This will remove the requirement for major capital investments to try out new ideas and concepts. We will see high school and college students building devices and tools that we never even knew we needed, bringing innovation to physical devices that they currently bring to software like Google, Facebook, Twitter and others."
Ian Verhappen agrees that wireless and digital communications will be critical to the future, but he cautions, "The impediment to adoption will be training and legacy systems."
Around the corner, he sees "smarter device integration with minimal end-user input and increased use of secondary, tertiary and implied measurements through real-time data mining and pattern recognition to not only tell us of reading abnormalities, but also process conditions implied by the combination of readings, rather than something read directly."
He continued, "We will also see more remote operations where the central control room will be a remote control room many miles form the process, much like is being done with offshore platforms used in locations that are difficult to reach or staff, such as oil sands facilities, underground mines and so on."
Dr. MacGregor says, "I have no doubt that the use of the massively increasing amounts of data that we are collecting on processes will be the most important factor influencing control engineering practice over the next few decades. How this data is used is the key to all control problems, and we are well beyond our limits already on being able to model all these measured effects theoretically. The data is also unlike data that most of statistics and control engineering is used to seeing in the past. The volumes of data are very large, and the variables are all highly correlated, thereby resulting in greatly reduced dimensional spaces. This is why I started up ProSensus in 2004 as a spin-off company out of the McMaster Advanced Control Consortium. It is a company that deals with all these problems of analysis, monitoring and control in the process industries."
Dr. Peter Martin said, "I believe that with the continued use of COTS [commercial off-the-shelf] technologies and development of industry standards, the technological details involved in automation and information systems will become less important and less differentiating over time. This has been happening for years, but I expect the trend to accelerate. But I expect the scope of what is considered to be an automation system to continue to expand significantly until single system domains encompass entire plants, enterprises and value chains to enable new value-generating solutions to be developed and delivered. I expect that much of the focus in automation going forward will be on the development of these new value-creating solutions."
He adds, "I believe the future will be a very exciting time for industry and for industrial automation. "When I joined the process industries back in the 1970s, I found it to be one of the most 'solution-oriented' fields I could find. Over the past 30 years, technological features overwhelmed much of this solutions-focus, and often these new technology features offered little or no value. I would much prefer to be in a solutions-focused environment in which the great minds of industry can concentrate on providing value-generating solutions that simultaneously improve the safety and environmental integrity of industry."
What We Do Matters
"You cannot have a sustainable society without the manufacture of goods," Martin says. "And we know how to manufacture goods profitably, safely and cleanly. Engineering may be the most valuable profession in industry and in the world. As an industry we need to communicate the value of what we do to society and get ourselves out on the forefront of this discussion."
He adds, "I believe I would do the same thing given the opportunity to start again. I believe that what we do is important to the world. We all want to have a positive impact on the world in which we live, and I believe our profession provides great opportunities to have that impact. Industrial automation systems are among the most sophisticated and high-impact technologies being used today. We need to communicate this. We also need to use the technology to provide more attractive work environments for all industrial personnel. People must regain their pride in making things that society needs. The automation industry can help make this happen."
On the same subject, Dennis Brandl adds, "I grew up with the TV show "McGyver" as a role model for engineers, but there is no equivalent today. As manufacturing professionals, we need to engage high school students and show them the fun involved in building robots, competing in science fairs and building model rockets and model planes. There will be some percentage that will find a lot of value in building 'stuff' and making 'stuff' work. These students need to see that a career in manufacturing can not only be fun, but it also can be profitable."
Verhappen agrees. "I think Dean Kamen is taking one right approach with FIRST Robotics, and the main reason for success is that this program reaches young people when they are making the decisions—middle school. We need to be able to come up with a way to overcome the concept that science and engineering are bad and show that good science and engineering, especially monitoring and control, are the best ways to help the environment while enjoying a good living. We also need to emphasize that university is not the only option, because we need at least as many instrument technicians as we do engineers, because once something is designed, someone needs to build and maintain it for the next 20-plus years."
MacGregor pointed out that students will opt for the fad choices and also for the most lucrative fields. "One good thing going for engineering in general is that the financial collapse has reduced the stream of people into those industries and made careers where you have to work for a living more attractive."
He continues, "In general, Canada has no problem, and Canadian universities graduate lots of process systems engineers, with the surplus moving to the United States. The problem is with the American universities and research funding agencies in the USA. The funding agencies fund the current fads (systems engineering not being one of them), and professors follow the money. The major U.S. research universities are mostly interested in training students to enter their graduate programs rather than industry, and they don't encourage a process systems engineering stream and, in fact, don't even offer control and statistics courses to their undergrads. So look to the U.S. funding agencies as the problem. The reason we have no shortage of people entering the process systems engineering stream in Canada is that the funding agencies do fund systems engineering, and it is considered a viable area of graduate research and a very attractive area in industry."
MacGregor sums up, "I have had a very rewarding career. I would tell my young relative that there will be some interesting and rewarding positions in the future in certain areas of process automation, particularly those that build upon the use of big data. This area will only demand more and more engineers in the future."