Automation IIoT Information Technology Management

Move From Raw Data to Smart Work

EP Editorial Staff | November 9, 2016

Manufacturers are flooded with data. Here’s some guidance to help you put that data into context, understand it, and make it work for you.

By Tim Sowell, Chief Architect, Software, Industry Solutions and Stan DeVries, Senior Director, Solutions Architecture, Schneider Electric

In today’s “flat world” of demand-driven supply, the need for agility is accelerating at a rapid rate. This is driving leading companies to transform their operational landscape (systems, assets, and culture) to a “smart work” environment. This move toward agility transforms thinking from a process-centric view to a product and production focus, requiring a dynamic, agile work environment between assets/machines, applications, and people. The paradigm shift from the traditional “lights-out manufacturing concept” of fully automated systems to an agile world of dynamically planned yet scheduled work requires:

• automated embedded intelligence and knowledge
• 
augmented intelligence using humans to address dynamic change.

1116schneider1

At the foundation of this shift is an environment in which a worker can have the mind space to understand the larger changing situation and make augmented, intelligent decisions and actions. To provide this, data are transformed naturally into operationalized information upon which decisions can be made, then combined with “tacit, applied knowledge,” providing incredible value when taking operational actions.

The explosion of information across industrial operations and enterprises creates a new challenge—how to find the “needles” of wisdom in the enormous “haystack” of information.

Listen to MT editorial director Gary L. Parr’s interview with one of the article’s authors.

One of the analogies for the value and type of information is a chain from data, through information and knowledge, to wisdom. In the industrial-manufacturing and processing context, it may be helpful to use the following definitions:

• Data: Raw information that varies in quality, structure, naming, type, and format.
• 
Information: Enhanced data that has better quality and asset structure and may have more useable naming, types, and formats.
• Knowledge: Information with useful operational context, such as proximity to targets and limits, batch records, historical and forecasted trends, alarm states, estimated useful life, and efficiency.
• 
Wisdom: Prescriptive advice and procedures to help achieve targets such as safety, health, environment, quality, schedule, throughput, efficiency, yields, and profits.

1116schneider2

To illustrate this transformation, imagine driving along an unfamiliar California freeway in a GPS-enabled rental car:

• Data: The GPS knows that I am on a freeway, traveling at 80 mph.
• 
Information: It is “situationally aware” that I am heading south on the I-405 freeway.
• 
Knowledge: It works with other services to determine that 10 miles ahead the traffic is stopped, and provides me with a warning that I will be delayed due to a traffic hazard. It has combined traffic knowledge with my location, speed, and destination to provide timely, advanced decision-support knowledge that I can use to potentially take an action.
• 
Wisdom: The GPS provides two alternate routes, giving me the time and characteristics of each route.

Without requiring me to take my eyes off the road and use an A-Z directory, I have been:

• warned ahead of time of an issue that could prevent me from reaching my destination on time
• 
given two alternatives and the information necessary to realize my goal with either choice.

There is no reason why this same transformational journey from raw data to wisdom cannot apply to manufacturing operations.

Avoid the Pitfalls

Many companies stand on the edge of a data swamp that is growing quickly, with the Industrial Internet of Things and Smart Manufacturing providing access to an exponential level of additional data from their industrial value chain. This data influx can either bog down growth or, if leveraged to achieve proportional knowledge and wisdom, create a new level of operational agility. The Fourth Industrial Revolution (Industrie 4.0) provides a framework for leading this ubiquitous transformation. Major industrial organizations are now realizing the incredible value that can be extracted from data and are combining time, resources, and technologies, such as big data and machine learning, with a new evolution in operational culture to leverage this potential.

Those operating in manufacturing have been living for decades with vast amounts of data located in historians, equipment logs, and across their extended supply-chain network. Data, in and of itself, is not of much value. The same can be said for reams of paperwork that document best practices—it isn’t of much value sitting on a desk or in a document-management system.

1116schneider3

Operational Data Management

We all talk about the ability to generate data from different devices. This can be valuable, provided there is some enterprise integration. But, can you really have effective information if there is no context?

The challenge is how to gain this context and then sustain it over several devices (things) without having a significant impact on those devices. In other words, how does one add, remove, and evolve devices? This requires an operational data-management system that is a “yellow pages” of the system, providing the context and relationship between devices and the operations.

An operational data-management system provides the ability to register new devices and data input, while maintaining the detail in the device, and then provide the bigger operational process alignment. This provides the association, which is alternate naming of that device so other applications can find and interact with it. Often, other systems and machines have a different outlook on the process and will use different naming and references for the device. An operational-data-management capability provides this association and ability to align many devices without a change in the underlying applications or devices.

From a data to information point of view, it provides the context needed to gather data and transform it into information, so that big-data analysis and other tools can be applied and convert that information into knowledge. Knowledge provides a pattern to ensure that contextualized operational data (production, quality, machine status) is integrated with templated collaboration activities, and ultimately broad value/supply-chain management.

Without this, companies have a real risk of gathering significant amounts of data and being unable to create the associated proportion of information, knowledge and, eventually, wisdom. Knowledge allows companies to put architectures and systems into place and gain contextualization while providing the plug-and-play ability for devices to be added to the solution.

The cost to store and share data has dropped significantly, and a simplistic expectation is that, although storage is growing by a factor of millions in only a few years, the pattern illustrated in Fig. 4 evolves.

1116schneider4

Although the pattern might seem to be logical, it is actually a nightmare, because it becomes much harder to discover and translate knowledge and wisdom from another operation, especially in another location, to the local needs. But there is a solution.

To understand the problem better, let’s consider that knowledge includes context. This context begins with local details, including time, location, process or machinery configuration, raw materials, energy, and products being processed or produced. It is already valuable to have wisdom to achieve and sustain best performance for the community, customers, and the corporation. This local context only needs to know its immediate information, if it has enough wisdom.

Now let’s consider what happens when a single site, a fleet of similar sites, or an enterprise have numerous similar operations. How can local wisdom be enhanced by using wisdom from other operations? Solving this problem is important for operations transformation, such as operating physical assets as one (in a chain or as peers) or by supporting multiple operations with a flexible team of remote experts.

One approach to solving the knowledge proliferation problem is to take advantage of a methodology used in distributed databases, where a technique called “federated information” is used. This technique is especially valuable in industrial operations-management architectures. Federated information does not change the local information’s naming or structure, but provides multiple translations, across the database, for multiple similar structures and for multiple contexts such as what financial, technical support, scheduling, quality, and other functions require. It is an alternative to the fragility and complexity of attempting to force a uniform and encompassing naming and structure in an attempt to satisfy all applications and users.

1116schneider5

The same approach can be applied for wisdom. Currently, hobbyists and enthusiasts around the world share wisdom for restoring cars, making furniture, playing a musical instrument, and gardening, as examples. Anyone with no experience at all can ask “Where do I get started?” and most respondents will provide some type of advice. In the same forum, experts can share wisdom that is valuable and understandable by them at their level of experience. This wisdom is extremely decentralized, and the experts are providing the local and regional translations.

In the industrial-operations environment, federating wisdom is partially automated by expanding the local context, including information about adjacent operations, about the chain or peers if these operations are being managed as one, and then knowledge is expanded by applying the context of group targets and performance.

Some enterprises have hundreds or as many as tens of thousands of similar operations supported by dozens or fewer experts. Discovery of wisdom is greatly enhanced by maintaining an architecture that enhances local context without modifying or attempting to force burdensome structures on local operations.

Empowerment through Wisdom

So how does operational intelligence/industrial analytics and the movement to wisdom relate? They are different, but all are related to empowerment of an operational workforce to make earlier decisions and take informed actions. One of the big drivers for platforms is to manage variance. We talk about supervisory, MES, information and simulation platforms but we also must have a people platform that covers:

• collaboration between people
• 
activity hosting, including embedded information/knowledge and associated actions
• 
transformation of information to situation awareness for the user who is interested/ interacting
• 
management of operational work between team members
• 
notifications.

This people platform will mitigate workforce turnover by abstracting the different skill and experience levels with embedded applied knowledge (wisdom), so the experience is now in the system. This is a key concept for operational transformation.

Industrial analytics provide the shift from the past through the present and into the future, based upon high-fidelity models gained from experience. It provides a new dimension to worker tools and transforms the decisions that they are about to make. Industrial analytics combine the future, providing answers to what will happen with the recommended actions to take.

This also provides the answer to “What should I do next?” with experience, forethought, and understanding. Operational intelligence furthers decision making by providing screens/presentations of the situation or known questions with context and awareness.

Operational intelligence provides the worker with an understanding of “now,” where he/she is, and what the future holds, simply and clearly. Increasingly, there is demand for this type of operational window and views. It is not analysis but practical information around the current situation and immediate future. Just a simple view of the task or question provides the clear awareness and actionable answers.

Are these different experiences? No, they are all functional value expansions on each other, and should be seen as building blocks on the road to providing an operational execution knowledge platform, with built-in experience. In other words, they provide a foundation for absorbing turnover and transition in the workforce while maintaining operational consistency and efficiency.

There is a journey to smart work that organizations are now following, much like the first continuous-improvement initiatives that began nearly 50 years ago, such as Lean, Six Sigma, and TQM. Operational execution puts it in the systems and culture that enable proportional growth in knowledge and wisdom so that they can address the dynamic world of smart work. The only difference is that operational data-driven systems can now be a part of these continuous-improvement strategies.

Manufacturing is in a constant drive to improve performance, and transformation of work has become the main method to achieve and sustain this. Higher capacity or more efficient machinery and processes aren’t sufficient anymore. Manufacturers with agile and cyclical operations need a method to remain cost competitive during the lower throughput periods, yet remain responsive enough to take full advantage of high throughput or high margin conditions.

Implementing systems that transform work using higher value information and reliability change when, where, and/or how users make decisions, and is the foundation for this next level of improvement.

Operational transformation through smart work is a journey, and technology is only one of the key elements. The user culture must adapt, similar to the previous waves of quality, safety, health, and environmental improvements. The journey advances with work-process improvements, as applied to sections of a site or an entire site. Existing software must be assessed in terms of delivering knowledge and wisdom and supporting mobile and traveling workers, with the goal of significantly reducing the skill and effort required to maintain them. The journey is worthwhile, practical, and essential for manufacturers not only to stay competitive, but also to thrive.

Tim Sowell is vice president of Software System Strategy at Schneider Electric, Lake Forest, CA. In this role, he leads the direction and strategy for the company’s Wonderware software portfolio. Stan DeVries is senior director, Solutions Architecture at Schneider Electric Software. He works with customers to implement innovative, reproducible data-architecture patterns and reference architectures.

FEATURED VIDEO

CURRENT ISSUE

View Comments

Sign up for insights, trends, & developments in
  • Machinery Solutions
  • Maintenance & Reliability Solutions
  • Energy Efficiency
Return to top