This week’s personal interview was conducted with long-time expert in industrial automation’s evolution, Ernest Roland. Ernest was the co-founder of Crisp Automation, as well as InduSoft.
Crisp Automation was an early SCADA system that was eventually sold to Atlantic Ridgefield in 1979. Ernest graduated from the Aeronautic Institute of Technology in Brazil, and has a Master of Science and Engineering from Ohio State, and an MBA from the Illinois Institute of Technology.
Here are some of our questions for Ernest, and a summary of his replies. You can listen to the full podcast here:
Q: You have a masters in electrical engineering, and a specialty in communication. What made you decide to move into Industrial Automation?
A: Well, we need to look at the environment of the mid 1960s, when digital computers were physically very large, power hungry, slow, and horribly expensive. They had competition from analog computers used for specialized engineering computation. Then came the integrated circuit, and it was a game changer. With it came logic boards and mini computers. Like Marcia mentioned, the PDP-7 had a 12bit word length, and 4K of memory, and could be acquired for the ‘bargain’ price of $20,000. It was still very expensive at the time, but it was cheap compared to alternatives. At that time, only very high value processes could afford digital automation, such as the million dollar IBM alternative used in many refineries and mills.
Q: And so with that price point, you saw a shift in the opportunity? Is that what led you to invest in the industry over other engineering sectors?
A: That’s correct. The minicomputer was a revolution. At one point before market consolidation there were 48 manufacturers of minicomputers. They were addressing the segment of automation below mainframes. The price point was still high, but by and large, much more affordable for other classes of processes.
Q: So you Co-founded CRISP Automation and InduSoft. What changes do you see in the market for process automation? InduSoft is purely a software company, while Crisp Automation did sell hardware also. What led to this dynamic?
A: Technical problems are always dynamic, with different solutions for different problems. For instance, when Crisp was founded in the 1970’s, the main game changer was the minicomputer. Applications grew exponentially, and most of them were used in continuous processes. We chose the niche segment of batch processing. We were dedicated to automating batch systems, because that’s a case where the computer can do a much better job than a human being. Humans are creative, but not repetitive enough for batch processing, and in certain chemical processes or in food and beverage, there is a lot of money riding on accuracy. Batch automation allows a more uniform result. That’s the environment we entered with Crisp Automation.
Q: So would you be surprised that today we still have people using Crisp?
A: No, I’m not surprised, because Crisp offered a unique way of doing batches that was very efficient, and is still very cost effective today. Of course now there are much better tools available and much better options for graphical interfaces, but the same principles of state diagrams, etc, are still a good way of thinking about batches.
Q: I saw an interview with Steve Rubin in Control Design Magazine in 2005 that said he left Foxboro to work at Crisp because he was enamored with the technology. What do you think made Crisp so unique? Was it just the Minicomputers, or was there something different that caused people to want to be there?
A: Well, first of all, we always had the lead on existing technology. If you don’t have something that is unique then there’s no point entering the market. Besides the minicomputer, we had chosen some equipment, and we developed some crude software tools for programming. We had proprietary software for batch processing which was very efficient, and we were the first company to use the internet in the industrial environment. At first, all the developers were creating proprietary networks, while we decided to use the very cost effective and reliable Internet. At the time, everyone thought it was non-deterministic and shouldn’t be used, but now no one questions it. There are always points of resistance created by new technologies, and in embracing the new technologies we were unique.
Q: With your history in Industrial Automation, what are the other biggest gamechangers that you’ve seen in your time in the industry?
A: Well the main one in the past 50 years has been the adoption of digital technology in computers and especially software. Industrial automation is a much smaller market than the IT market, and the trends developed in IT have slowly migrated to industrial automation. Industrial automation has some particular constraints, such as where equipment is installed, whether or not the environment is harsh, or gloves are needed, etc. But in essence, the basic technology is always borrowed from IT. The major change I saw was the power of software.
Originally we offered whole solutions, and believe it or not, we gave the software away. Now at InduSoft, we only sell software, because I/O hardware became a commodity; there’s no need for only one reseller. With software, the Microsoft and other operating systems like Linux and VxWorks mean that software can be used on a variety of different hardware options. But during the time of the minicomputer, with 48 different manufacturers, everyone had a different software approach.
Q: Now we are seeing the industrial world adopting other trends from IT, such as merging OT and IT, and industrial automation together. Do you agree?
A: I see it slightly differently. IoT is slightly different from industrial automation or Industry 4.0. In IoT, the applications are designed to get simple physical data to bring it to the average user. But we still have specialized needs for the plants, and the typical plant engineer needs to be convinced to join the current IT world. If they could operate independently, they would prefer to. Therefore, what we see is more I/O being installed that is not necessarily for control of the process, but for the dollars and cents side of production management.
Q: What changes and trends came as the biggest surprise in your career?
A: The biggest surprise was the very rapid adoption of smartphones and mobile devices. They’ve really taken off, and it’s good that InduSoft has a good solution for smartphones and tablets.
Q: What advice would you give to someone entering the field now?
A: Look carefully, and chose technology that won’t lock you into a corner. Don’t fall into a position of maintaining an older, proprietary system that will become obsolete. It will make your knowledge of little value. Always remain aware of the changes in your field. If you’re an entrepreneur, be it a system integrator or a developer; understand that you’re living on a high wire. Don’t aim to be ahead of technology – Bill Gates is a great example. He was never exactly ahead, but he was ready to adapt to technological changes as they arose. If you move too far ahead, you run the risk of being the only one supporting a standard or a technology that isn’t widely adopted. You also can’t call behind, because then you become quickly obsolete. Markets disappear in industrial automation just like Kodak film has disappeared.
Q: Hindsight is 20/20. If you had to give yourself advice 10-20 years ago, what would you have done differently?
A: Many things! But more important is what I would not have done differently, which is the people I chose to work with. When you enter the market, you need to be sure of your business model. If I had to do it differently, I would not have done so much switching around in earlier years. I would have concentrated more on my business model and stuck to it more closely. In the earlier years of Crisp, with 48 different manufacturers, it was very tempting to jump from one to another and go into different markets, because we didn’t have a generalized solution. So the advice I would give myself is “Keep the same great staff, but concentrate!”
Q: Today everything is focused on remote monitoring and online connectivity. When did you begin noticing the switch to remote and online communication?
A: It always existed, but it wasn’t always adopted because you need the enabling technology, and that didn’t always exist. But from very early years, the question has always been economic. In some applications such as water treatment or drilling operations, you need some way to monitor what is happening in remote locations. There were many companies that had specialized solutions for pipelines and waste treatment plants. But now the solutions are more generic, and can accommodate remote monitoring even if it’s not vital to the process.
Q: As we move forward, what role do you see the Internet of Things playing in the larger automation industry?
A: Well, just like motors used to be so expensive you’d only have one or two in a mill and now we have dozens in our homes we don’t even know about, I think there will be hundreds of embedded processors. I see that being the biggest new frontier. What will we get from all this data? But I see the progression as unstoppable. Not everything will take off, but some kinds of data collection devices will be ubiquitous.
Q: How do you predict industrial automation will change in the next ten years?
A: I think we just said it. The biggest thing in the future will be analytics. We will have displays that are much more human oriented to help us understand this data, and how much is automated will change. Much of the data will pass through mobile devices. The industry will be modernized.
Q: Do you think that’s what HMI/SCADA customers expect – modernization?
A: Yes, but I think that the term HMI/SCADA will not be as widely used, because the market is evolving into a set of very smart tools that do the jobs of HMI, SCADA, and even some more. There will be end-point solutions for specific problems, and I see this as the big next wave, using the incredible amount of software that’s being generated worldwide.
Q: Do you have anything to add before we conclude?
A: We’ll soon be creating much better interfaces. The current generation playing video games will be the ones creating and programming future interfaces, and they’ll bring their own experiences to the table.