Boundless Automation Ambitions

ARC met up with Peter Zornio, Chief Technology Officer, Emerson, during his recent visit to the company’s Asia Pacific headquarters in Singapore. Among the highlights in a wide-ranging discussion, which touched on everything from corporate strategy to plant field networks: the move for Emerson to become a pure-play automation company; the logic and landscape for acquisitions; the progress of the AspenTech partnership and integration; and the vision and concept of boundless automation.   

ARC: Since we last met and spoke in Singapore pre-Covid, there have been quite a number of changes corporate-wise at Emerson. Can you explain the rationale behind these?

Lal Karsanbhai, who was heading up Automation Solutions, was promoted in February 2021 to the position of president and CEO of Emerson, i.e. the whole company. At the time, in addition to Automation Solutions, which was about two-thirds of the business, we had the Commercial and Residential platform, centered mainly around HVAC and climate controls offerings, contributing the other third. Lal really believed that we should double down on automation and become a pure-play automation company. And he's been executing that strategy over the last couple of years. 

ARC: So what does Emerson look like now?

Last October we announced that we were selling a majority (60 percent) of Climate Technologies, the largest piece of Commercial and Residential, to the private equity company Blackstone. We did keep our tools business, which has been renamed Safety and Productivity, but that's a small part revenue-wise of Emerson. (ARC note: the acquisition by Blackstone was completed on May 31 with Climate Technologies valued at $14 billion.)

As a consequence, Emerson is indeed now very much largely automation. What that means structurally is that the layer of Automation Solutions organization between Corporate and business units like Measurement Solutions and Flow Controls has gone away. So for me, for example, whereas I was previously CTO of Emerson Automation Solutions, I am now CTO of Emerson. I do have some expanded responsibilities, including managing our big development center in Pune, India, where we have around 1000 people working on technologies across all of our business units.    

ARC: With the focus on automation and the company replete with cash, acquisitions must be high on the agenda?

Well, obviously, we would love to buy more automation companies. But there are not as many easy targets, at least right now, as you would think. Because if you really look across the world of automation, both process and discrete, you've got the big players, which we're clearly not going to buy, and then you have a lot of smaller companies that are privately held and not for sale, especially in factory automation. It’s been a few years now since we bought GE Intelligent Platforms, which gave us industrial PCs and PLCs, and that business is now part of what we call our Discrete Automation Group along with pneumatics technology brands like ASCO and AVENTICS. Growing the discrete automation side of the business is one of our strategic objectives.

ARC: The offer to acquire to National Instruments took many people (including me) by surprise…

Well, as Lal articulated at the 2022 Investor Conference, for acquisitions, we are looking for companies in very adjacent spaces to automation. And one of the spaces that came back in our analysis as an attractive target was the test and measurement world, which has similar technology, in terms of sensors, hardware and software, to the automation world, and is definitely much closer to what we do than Commercial and Residential was. 

The other big attraction of National Instruments is its faster growing markets than we typically serve in automation. Markets like electric vehicles, for example, which is a big business for NI given the rising demand for test equipment and solutions for batteries for EVs, and driver assist systems in all vehicles.  Semiconductors is another; money is pouring into that space right now as countries look to onshore and build infrastructure in the light of the geopolitical concerns around China. (ARC note: on April 12, a definitive agreement was announced for Emerson to acquire National Instruments for an equity value of $8.2 billion.) 

ARC: The acquisition (55% majority share) of AspenTech completed about a year ago. How is the partnership coming along?

With very few exceptions, our portfolios were very complementary, which set the foundation for a good partnership. They had nothing in terms of control systems or instrumentation, and while we were starting to push into AspenTech space somewhat with Plantweb Optics Data Lake, we had nothing in process modeling to the extent of Aspen Plus or HYSYS, which we now use as the core software for our OTS (operating training simulator) solutions. 

So, the one area where there was a little bit of overlap was data management and analytics. Well, as soon as the deal was announced, we recommended they buy a German industrial information management company, inmation, we were working with and in which we owned a minority share. Last October they did just that, and have put it into a new business unit, AspenTech DataWorks, for managing industrial data. The Plantweb Optics Data Lake name will go away, and all our customers who had that will now become AspenTech customers. 

ARC: And presumably, Emerson is still looking to grow in industrial software?

Absolutely. Industrial software is a targeted strategic growth platform along with discrete automation and test & measurement. But what’s different with industrial software is that the acquisition vehicle will be AspenTech and any buys in this space will go into the AspenTech organization rather than into Emerson. The reason for that is quite simple: software companies trade at much higher EBIT multiples than traditional automation companies. So that’s why software companies that we purchased previous to AspenTech like Open Systems International (OSI), which provides solutions for power grid management, and Paradigm, supplier of software for subsurface oil and gas, now sit within AspenTech. More recently, Australian mining software company Micromine will go into AspenTech when that acquisition closes. 

ARC: We have to talk about boundless automation…

Yes, this was a big announcement at Emerson Exchange last October, where we introduced the term and the concept. Boundless automation is essentially our vision of the future automation technology architecture. It recognizes the limitations of how automation is deployed in plants today, with purpose-built hardware and the hierarchical layers of sensing, control and information, which inevitably creates data silos, limits free and secure movement of data, and separates IT from OT. 

The new automation architecture will be software-defined, data-centric and inherently secure by design. Software applications will execute in three computing domains: intelligent field devices; edge computing/control environments; and in the cloud. Unifying these domains through a cohesive software platform will democratize data, make it much easier to access and transfer to value-adding applications and analytics tools. 

Boundless automation is not going to happen overnight; this is much more a multi-year kind of journey. But we felt it important to put a stake in the ground about what we believe is going to happen in the future.

ARC: What are going to be the implications for control systems like DeltaV?

Software for process control that is tied to purpose-built hardware is eventually going to go away ‒ for everybody. Executing control loop containers on a highly available hypervisor will be able to deliver the kind of latency and performance that we get from hardware specific controllers today. Decoupling hardware and software is a key aspect of the boundless automation vision.  

ARC: Dare I say this sounds a lot like the OPAF (Open Process Automation Forum) initiative?

There is some commonality in that we both believe that control system software will no longer be tied to specific hardware as it is today for all DCS vendors. Where we would fundamentally disagree is that while the OPAF standards may provide a minimum level of compliance, they will not be rich enough to provide the level of unification and integration between the automation software applications that people want. You have, say, 100 different vendors making 100 different pieces of software and a systems integrator puts any combination of them together and they all work and play with each other the same way the software in a DCS does that was all designed and written by one vendor? 

They tried something similar in the ERP world in the 1990s with middleware ‒ take the best of breed  accounts payable software, the best HR software, etc. from different suppliers and put them together. Most companies gave up. What did they go to? Integrated suites of software from a single ERP supplier, which is what you have today from Oracle and SAP, although even as a single vendor they still struggle to truly make it unified.

OPA is a great vision. Everybody loves the vision. Who wouldn’t want to buy this software from one supplier and have it work seamlessly and transparently, with no effort, with another piece of software from another other supplier? But the level of work it takes to do that in reality is immense. Every time any one of those pieces of software has a security upgrade and stops working with the rest of the software, or a new release comes out and it stops working in the rest of the software, then who is going to be responsible?

We believe in having a software ecosystem with one automation vendor infrastructure, not a systems integrator trying to tie together software from multiple different vendors. In our vision of boundless automation, the unified suite of software will come from us and partners, not from just any third party. We will be in charge of our own destiny. Of course, we will still support all the open standards and allow other 3rd parties to integrate in increasingly richer manner; but the ease of use and support you get from truly unified software requires someone to have overall responsibility.  Apple has certainly proven consumers like the ease of use, security, and single responsibility of a unified portfolio.

ARC: Given the extent of the process instrumentation portfolio, the new Ethernet Advanced Physical Layer (Ethernet-APL) should be very significant for Emerson?

We are working with OPC Foundation and FieldComm Group committees on Ethernet-APL and engaged very much in the technology that will enable the use of physical Ethernet cable in process environments, including in hazardous locations. The benefits of deploying APL are going to vary greatly depending on the complexity of the device; the more data that's in the device, the more it needs the extra bandwidth and power that APL can provide. 

So Ethernet-APL is highly differentiated for Emerson products like Coriolis flowmeters, radar level gauges, valve positioners and gas chromatographs But for pressure and the temperature transmitters, not as necessary, because with a couple hundred or less parameters in the device, they don't need all that 10 Mbs of bandwidth you get with Ethernet. Right now, I am coordinating a development roadmap across the businesses and making sure APL is high on the agenda for the businesses where it needs to be a priority. 

What Ethernet-APL has going for it that FOUNDATION fieldbus did not is that it’s just a new physical layer. So, for people who are already familiar with Industrial Ethernet protocols, that part doesn't change. Because the vast majority of process instruments sold today are HART devices, the simplest and easiest protocol for everybody is going to be HART-IP. The device itself will look exactly like the HART device they configure today, except it will be 8000 times faster with Ethernet-APL.