Emerson busts barriers with Boundless Automation
The aim of process automation is the same as ever—see, decide, and act. The difference is that Emerson wants to reduce or eliminate all the old layers and silos that keep getting in the way, and make these basic tasks much easier to perform.
The company reports this watershed improvement can be accomplished by replacing the seven-layer Purdue open system interconnection (OSI) model (ANSI/ISA-95) that’s defined how process control systems have been organized for several decades with a software-based architecture that democratizes data and enables optimization without all the former hurdles. This new “Boundless Automation” strategy connects operations technology (OT) and information technology (IT) domains, and moves data seamlessly and securely across them from field and edge devices to cloud-computing services and back to greatly enhance operations and enterprise performance.
A trio of Emerson’s experts detailed its vison in a media briefing on the first day of Emerson Exchange 2022 this week in Grapevine, Texas.
Demolish the layers and silos
To achieve its boundless vision, Emerson CTO Peter Zornio reported that end users and their supporters must first address the silos that have sliced and diced process automation for many years, creating pockets of data and enterprise. For example, it’s presently subdivided horizontally by field devices such as control and safety systems and other devices; onsite controllers and support components; site operations such as operator and engineering applications, historians, diagnostics and other software tools; and enterprise software that runs in data centers or the cloud, and stores information in OT data lakes.
“It’s been said that architecture follows organization, so we have the Purdue model’s traditional automation and software layers. However, this purpose-built automation creates hierarchical automation networks, including hard segmentation between OT and IT, complex security constructs between layers and devices, and hardware designed for specific layers,” said Zornio. “It also prompts OT people to say to IT, ‘Here’s your data for the enterprise. Now go away because you don’t care about uptime.’ However, this response isn’t enough anymore.”
Likewise, Zornio added it’s also assumed that data always flows one way—up from the sensors on the plant floor to the boardroom and the rest of the enterprise. And, it’s assumed these upper layers managed by IT will somehow insulate and protect the operations layers below. The silos and the prejudices they maintain impede efforts to access and use data universally. “The traditional solution is to just add another layer of data aggregation and create a data lake, but this isn’t enough either,” explained Zornio. “There has to be a better way to structure information, so it can be used across enterprises and looked at by different applications.”
This was a primary motivation for Emerson’s recent acquisition of Aspen Technologies and its software solutions, including AspenTech’s own purchase of Germany-based inmation and its data management software. Both software portfolios are expected to enhance the capabilities of Emerson’s longstanding Plantweb digital ecosystem, and pave the way to Boundless Automation and the benefits it can provide.
Unbounded architecture aspects
Instead of cementing in place the layers and silos of the past, Boundless Automation consists of three flexible computing domains—Intelligent Field, Cloud and Edge—which seamlessly interact around a Cohesive Software Platform. This environment has common, consistent data models and application program interfaces (APIs) that provide ease of use, data access among domains and a zero-trust security architecture that deliveries dynamic functional zones of authorization. In addition, Boundless Automation has functional systems for control, reliability and other tasks that manifest as software suites distributed across the common computing infrastructure.
The basic sections of Boundless Automation are highlighted by:
- The Intelligent Field includes devices connected to physical assets and measuring or controlling elements of those assets that are capable of digital connection. Its smart devices continue to expand into domains other than automation and add applications and computing power.
- Similarly, the Edge includes computing devices hosted close to where data is generated, and usually where data hasn’t yet traversed the public Internet. It securely places computing power closest to where it’s needed.
- As elsewhere, the Cloud includes remote banks of computing capability with an integrated compute fabric to host workloads that are typically from external organizations. It delivers infinite scalability, global connectivity, attractive lifecycle costs and rich analytics capabilities.
- At their center, the Cohesive Software Platform includes a unified data model, published APIs, multi-vendor apps, standards and secure-by-design capabilities. It’s also integrated by design and provides ease of use, data democratization, solution-level security certification, unrestricted innovation, interoperability, extensibility and unlimited scalability.
“Connectivity is important, but it doesn’t tell users what their data means or give it context. What they really don’t want is another layer or model. What they do want is this inherent computing environment that’s integrated by design,” said Zornio. “Other platforms like this include Oracle and SAP, though the one with the most ease of use is likely Apple’s ecosystem. Such a cohesive software platform democratizes data so users will know what their information means consistently across applications.”
Out in the Intelligent Field
Laura Schafer, VP and GM of Pervasive Sensing and connectivity at Emerson, reported that the Intelligent Field aspect of Boundless Automation has many familiar connectivity options for relaying sensor data and other plant-floor information. These choices and protocols include IO-Link, Ethernet Advanced Physical Layer (APL), HART, Wireless HART, and wide area wireless networking like 4G LTE and 5G.
“Wireless devices can drop in data for pervasive monitoring, either by using stationary, high-density industrial Internet of Things (IIoT) from Wireless HART, or in the form of mobile, single assets equipped with the 5G that’s moving into discrete and hybrid manufacturing,” said Schafer. “We’re also working on Ethernet-APL, which is unlocking the ability to use Ethernet in hazardous areas, and allows more data to safely flow from the surface up through firewalls. Likewise, WirelessHART is still ideal for safety because of its superior reliability and typically longer battery life, which allow users to ping it and get an immediate response.”
To make use of Ethernet-APL in the field, Schafer added that Emerson recently launched its small, compact and lightweight AMS Wireless vibration monitor that transmits data to the cloud or the edge. On the mobile, single-asset side, it’s demonstrating at Emerson Exchange its 408 Autonomous level sensor for remote tanks, which can measure and send level data to onsite users or cloud services. The company is also exhibiting an APL switch that’s enabled its CHARMS electronic marshalling I/O to take advantages of Ethernet-APL. “It can be dropped in next to legacy devices, so users can check it out where it makes sense without ripping and replacing,” added Schafer.
Emerson is also using the Process Automation Device Information Model (PA-DIM) unified data model to enable easier integration and apply greater intelligence. For instance, it’s demonstrating at the event how PA-DIM and increased intelligence in field devices can assist valve condition monitoring, achieve approximately zero unplanned downtime for critical valves and reduce major maintenance event costs by 50%. It’s also showing how smart field devices can optimize compressed air, saving $80,000 per year in maintenance and reducing compressed air consumption by 20-30%.
Forecast for IT? Cloudy
Once data comes in from the field, Zornio explained the Cloud side of Emerson’s Boundless Automation vision parallels how cloud-computing is taking over IT elsewhere, and gaining an increasing role in OT.
“Initial skepticism about the cloud is dissipating rapidly due to digital transformation programs that are pushing OT and IIoT together,” said Zornio. “Cloud vendors are adding services, so applications and users can ingest data and power their analytics in domain-independent ways, such as using Microsoft’s Power BI software. The cloud has a lot of appeal. IT loves it, and OT will, too. This is because it allows enterprise-level coverage, capabilities and access. The cloud also brings a whole additional paradigm of tailored, on-demand support, such as we have with our MyEmerson program.”
Zornio added that the cloud is even revolutionizing how software is developed. This is exemplified by today’s well-known software containers, which are sets of functions and services that can run almost anywhere, easily communicate with each other, and maintain security. Similarly, fleet management software orchestrates provisioning and upgrades entire software fleets, while multi-tenant capabilities reduce execution costs and allow single-instance upgrades. Finally, software platform-as-a-service (PaaS) accelerates app development and outsources software infrastructure management.
Back to the Edge again
Similar to the field, but with computing, the Edge aspect of Boundless Automation offers a new way to deploy software workloads and connections, according to Claudio Fayad, VP of technology, process systems and solutions at Emerson.
“We can’t talk about Boundless Automation without the Edge. It leverages cloud technologies, but it’s also built on present OT infrastructure and what we’ve already been doing for 30-40 years,” said Fayad. “It also deploys software as orchestrated workloads, redefines the communication architecture, is data-centric, and is secure-by-design for each node. The Edge is where IT and OT innovations can come together, and where we break all the silos.”
For example, Emerson and AspenTech are enabling the evolution of the industrial edge with:
- Distributed compute nodes for control, safety, energy transportation, discrete and machine control, and machinery health and protection
- Smart fieldbus protocols like HART, Foundation Fieldbus, Profibus and OPC UA
- Cohesive software environment for engineering and design, control and optimization software, production management, and asset reliability and performance
“The edge is really a collection of hardware running on one operating system (OS), so everything can act as one environment,” explained Fayad. “It’s an automation environment that erases the Purdue model’s functional silos, and instead converges embedded functions as software-defined workloads. This eliminate bottlenecks, combines OT and IT protocols, adapts to required latency, lets workloads execute in the most effective locations, and increases data accessibility.”
In fact, Fayad added, Emerson has incorporated edge automation in the recent release of its DeltaV, Version 15, distributed control system. DeltaV Edge Environment will be available in early 2023. It features a secure replica of DeltaV data and context, a secure sandbox to run analytics and custom code, and connects with the cloud and other edge workloads. The company also unveiled its PACedge that consists of an edge controller and IIoT software, which integrated real-time and non-critical workloads, and connect former islands of automation.
Because data no longer needs to go through so many firewalls and layers, Fayad reported that the edge makes communications more flexible and simpler, but still makes them secure-by-design. “We can put the protocols for the right job where we need them, and move data in bulk with its context to where it will be consumed,” added Fayad. “Once all these terabytes of data come together, they can be integrated, contextualized with data models and democratized.
“For instance, I can have a unified data model for my individual pump or other device. With the silos of the past, if I wanted to know what happened at my plant today, I’d have to call five or 10 people, and take 10 days to find a root cause. With a unified model, the data for a device doesn’t belong to one person but can instead be shared with to whoever needs it, and quickly show what happened at the plant.”