How 1995 Shaped the Future of Automation
Published on : Saturday 06-11-2021
1995 was a turning point that marked the end of an era and the beginning of several new technology paths, says David W Humphrey.
In the world of industrial automation, nothing ever happens fast. We don’t know revolution, only evolution, and there’s a good reason for that. But we do mark milestones every now and then, like the beginning of Industrie 4.0 in 2011. However, for those of us who have been around long enough, 1995 was a special year because it marked the end of an era and the beginning of several new technology paths that our industry has followed for more than a quarter century.
What ended in 1995? The era of big, expensive PLCs. The use of antiquated and sometimes proprietary systems for visualisation. Slow, vendor-specific industrial networks.
What was new in 1995? Plenty!
1995 marked the advent of modern PLC architectures
The year 1995 witnessed the launch of two new PLC systems from Siemens (SIMATIC S7) and Rockwell Automation (ControlLogix). Both systems were a huge departure from the heavy metal that the industry had been accustomed to since the 1970s. These systems were much more modular and compact than their predecessors. They also looked a lot better – both showed the signature of experienced industrial designers who convinced the world that controllers should be aesthetically attractive products, even if they spent most of their time locked in electrical cabinets.
Most importantly, these controllers were more than just PLCs. Several years later, ARC coined the term programmable automation controllers (PAC) to describe them as platforms that support multiple automation tasks rather than just logic control. The all-in-one controller was made possible by fast, intelligent backplanes that greatly improved the communication between multiple controller modules, special function modules, and IO. In fact, microprocessors were added to “normal” analog and digital IO modules to make them intelligent, software configurable, and capable of reporting diagnostic information.
With the new controllers came new engineering tools that made automation configuration much simpler. Tag databases were introduced that allowed users to give IO points meaningful names rather than cryptic addresses. Intelligent modules, such as high speed counters or weigh-scale modules could be configured directly via the programming software, allowing data to be scaled in the device first, then used directly in the PLC program in meaningful units. This led to the introduction of “engineering frameworks” that integrated programming and configuration in a single engineering tool. Probably the greatest advancement was the integration of motion and logic control. Suddenly, complex, coordinated motion among multiple servo motors could be programmed directly in the PLC rather than in another language and another module.
1995 witnessed the birth (and burial) of the soft PLC
In 1995, we learned that PLCs could run on PCs, thus launching the era of PC-based control. The PLC would be delivered on a CD and would run on a PC operating system as a protected task to ensure consistent performance. Data exchange between the PLC task and other software running on the same PC would be greatly simplified. Data logging would take place on the same machine with unlimited capacity. IO could be connected via network interfaces inserted into the back of the PC.
Would we replace our PLC hardware with much cheaper PCs? For the most part, this never happened.
The soft PLC was developed and marketed by a few small, innovative companies, but these firms were quickly acquired by Big PLC where the products languished for decades. Some tried to package soft PLCs with ruggedised PC hardware, while others simply made the PC hardware look like a PLC, helping to blur the line between “hard” PLCs and PC-based automation. Now, a quarter century later, the soft PLC may finally have its day, but the vision has developed. Today, users can virtualise soft PLCs and consolidate them on servers either on-premise or, someday, hosted in the cloud.
1995 taught us to think and work in Windows
The magical year 1995 saw the introduction of Windows 95, the first iteration of Microsoft Windows that defined the look and feel of operating systems through today. Visualisation systems before that ran on MS-DOS or obscure operating systems. Configuration tools were primitive and complex, and graphical interfaces were anything but standardised. Windows 95 inspired the industry to abandon the heavy, proprietary hardware in use and to embrace the modern PC architecture as the platform for visualisation in the plant.
But Windows 95 and its successors were far from perfect. Most notably, manufacturing companies quickly realised the uncertainty and the sheer danger of managing software updates on factory floor systems as if they were office PCs. Software lifecycles were much shorter than those of expensive industrial equipment, and despite the growing footprint of its operating systems in industrial applications, Microsoft remained tone-deaf to the long-term needs and priorities of industrial users. Despite this, industry users quickly grew accustomed to doing everything in a windows environment, and even truly open operating systems like Linux never had a chance.
1995 saw the development of useful standards like ISA-95
By the 1990s, manufacturers were embracing the power and flexibility of spreadsheets to support the daily tasks of production management. The problem was, of course, that each “solution” was unique and therefore limited. Software vendors recognised a new market for solutions that would automate production management. Fortunately for the industry, work began early to standard definitions and nomenclature of program management and the way it interfaces to ERP systems, resulting in the creation of ISA-95 in 1995. The new standard created consistent terminology, information and operations models that are the foundation for clarifying application functionality and how information is to be used. Not all standards are created equal, and ISA-95 stands out as a truly useful standard that shaped future development in the growing market of plant software.
1995 was the year industrial networking missed the boat
Actually, this trend began a few years before in the early 1990s. This was the time when large suppliers introduced their own industrial networks like Profibus and DeviceNet, and then graciously decided to give them to the world for free as “open” standards. While the move seemed benevolent, it retained a certain advantage to the suppliers behind the networks, a strategy that ARC called out as pushing “open-proprietary” networks.
The boat that our industry missed was Ethernet. While the commercial world had settled on Ethernet as the de facto standard by the mid-1980s, automation suppliers decided that Ethernet’s lack of built-in determinism disqualified it for use in the factory. Our industry wasted another decade before realising that Ethernet was indeed the way forward, but only if suppliers modified it with non-standard technology to make it suitable for real-time control. The result was the roughly half dozen flavours of “industrial Ethernet” in common use today. Today, a quarter century after this missed opportunity, automation suppliers are working with IEEE committees to add a missing timing mechanism (TSN) and a common protocol (OPC UA) that may someday lead to a truly universal industrial network that adheres to the Ethernet standard.
Conclusion: What did we learn from 1995?
1. The most important lesson we learned was to embrace technologies from the commercial world rather than invoking the “not invented here” excuse. Now, in 2021, we are facing a similar situation. The influence of IT on OT (operations technology, aka “automation”) has never been stronger. Architectures are evolving with the simultaneous emergence of the industrial edge and the cloud. Hardware is disappearing as more hardware functions become “software-defined,” helping users to simplify architectures while adding flexibility. Proprietary hardware in industrial solutions is threatened by the inevitability of commoditisation.
2. In 1995, we learned to treat software as a valuable industrial asset. Driven by the vision of Industrie 4.0, industrial users are turning to software to extract hidden data from their processes and turn them into useful actions. While off-the-shelf solutions are typically the first choice, more and more users are creating their own custom apps, empowered by low code/no code tools that allow fast and efficient development and deployment. As the portion of software in industrial solutions grows, users are learning the importance of managing software just like any other asset.
3. Finally, 1995 taught us the importance of embracing industry standards. Rather than ignoring established standards or bending them back into proprietary technology (see “industrial Ethernet”), our industry has learned to take a seat at the standards table early in the game and ensure that interests and priorities of industrial users are represented in new technologies like 5G.
Article courtesy: ARC Advisory Group
ARC Insights are published and copyrighted by ARC Advisory Group. The information is proprietary to ARC and no part may be reproduced without prior permission from ARC.
David W Humphrey, Director of Research, ARC Advisory Group, Europe. David is a member of the analyst team at ARC Europe and is based in Munich, Germany. In addition, he is a member of ARC’s hybrid manufacturing, packaging, and industrial networking teams. David has over 25 years of experience in industrial automation, including specifying, designing, and programming control systems in areas ranging from automobile to packaging, implementing projects involving PLCs, HMI hardware and software, industrial networks, drives and motion control. For further information or to provide feedback on this Insight, please contact the author at [email protected]
