A New Digitalization Strategy Framework to Advance Reliability and Asset Management

An Internet of Things Knowledge Domain Creates Stakeholder Alignment and Common Language

Sensors, cloud computing, artificial intelligence (AI) and machine learning – the Internet of Things (IoT) is coming to your plant or site! The IoT is expected to Revolutionize how you work. You likely have high hopes of its potential, but you also may have concerns about the volume, velocity, variety and veracity of data the IoT will bring and how on earth you will manage it. Security, the threat of malware and data integrity also may be weighing on your mind. So, you need a digitalization strategy, a plan and a framework for your success in an IoT enabled organization. 

A Common Language Approach

Today, organizations have a solid reliability and asset management framework in the Uptime® Elements released by Reliabilityweb.com in December 2013. This framework helps achieve a common language and system for organizations who understand that reliability is cross-functional and requires an enterprise approach.

Figure 1: The Uptime Elements Framework for reliability and asset management.

In Internet time, December 2013 seems a lifetime ago and although the IoT and digitalization strategies were born, they were not mainstream. Today, they are advancing like a bullet train and are now ubiquitous in all industries. 

A year ago, a small group of Uptime Elements enthusiasts, who also have considerable knowledge and experience, began a project to extend the original Uptime framework into the digitalization realm. 

A Word About Digitization Versus Digitalization

Words matter, which is why for this article, the word digitalization, as opposed to digitization, was carefully chosen because it is evident entirely new business models are emerging from this context.

According to Gartner’s IT Glossary, “Digitization is the process of changing from analog to digital form. Digitalization is the use of digital technologies to change a business model and provide new revenue and value-producing opportunities. It is the process of moving to a digital business. Digital business is the creation of new business designs by blurring the digital and physical world.”

As organizations move to implement the IoT and digitalization, people’s jobs will change. Imagine factory workers exchanging traditional handheld vibration analyzers, portable leak detectors and grease guns for machine learning algorithms, informed dashboards and programmed collaborative robots (cobots).

Digitalization Drivers

Based on a study conducted by Reliabilityweb.com in 2017 and 2018, the top three outcomes reliability and asset managers sought from implementing a digitalization strategy were:

1. Increased Reliability;

2. Make Better Decisions;

3. Decrease Cost.

While these are positive drivers, the research also indicates the pace of digitalization diffusion is increasing significantly within the reliability community. Most organizations lack formal policies, strategies and plans to align these implementations to organizational objectives or aims. Equally, conflicting viewpoints on the interpretation of what defines digitalization and how it applies to organizations make it challenging for companies to plan where and how to start. While some companies have hesitated to tackle digitalization, others have gone all-in without understanding how value will be delivered. These companies may find that the only business victory they achieve is bragging rights that they have implemented a wide array of science projects.

Thus the need for a unified framework around digitalization, one that gives common terms and an understanding of the challenge areas so they can be communicated across organizations and industries, demystifying the topic. Your enterprise, undoubtedly, will have different operational needs and IT architectures. You will need to assess what areas and guidance of the framework provide value to you. It is not a mandate, but instead a baseline of common terms, good practices and guidance to assist organizations that are looking to make the digitalization journey.

The Digitalization Strategy Framework

The Uptime Elements Internet of Things Knowledge Domain is a digitalization strategy framework to guide you in your reliability and asset management journey. It includes the Industrial Internet of Things (IIoT), cloud computing and more. Industry 4.0, as it has been named, now must be included in your road map, so this framework contains questions you should ask your organization and provides answers to common questions to clarify what you need to consider in implementing the IoT.

The elements within the IoT knowledge domain are:

  • SOURCE – items that generate or are sources of data;
  • CONNECT – methods of exchanging data;
  • COLLECT – preparation and storage of data;
  • ANALYZE – conversion of data into insights;
  • DO – actions taken from the insights.

Figure 2: Reliabilityweb.com asset manager survey on IoT, results 2017-2018 

Figure 3: The Uptime Elements IoT knowledge domain and digitalization strategy framework

Regardless of which element you are implementing, you need to consider:

  • The role of the digital twin, along with your automation and business systems in the context of the IoT: It is conceivable that your future state will be to conduct your work from the digital twin of your asset to better visualize and understand information in context within an accurate detailed digital model.
  • Trustworthiness of your data and systems: Identify and mitigate risks from security breaches and malware, and address privacy, reliability, resilience and safety.
  • Standards and governance: Learn from IoT data governance guidelines and configuration management principles to ensure data can be trusted.

Let’s take a more detailed look into each of the elements within the IoT knowledge domain.

SOURCE (Sc) 

ITEMS THAT GENERATE OR ARE A SOURCE OF DATA

Simply said, a source of data is the originating point of information used to inform situational conditions of people, places, or things. Sources can come from physical sensors attached to the things they monitor. Sources also can come from other external systems, such as social media, internal business systems (e.g., enterprise asset management system, manufacturing execution system, or enterprise resource planning system), as well as archives known under common names, such as data lakes.

Common types of physical sensors include devices that monitor things like temperature, proximity, pressure, water quality, chemical content, gases (present or not), smoke, infrared, fluid level, images (e.g., optical, thermo, infrared), motion detection, accelerometer, gyroscope and humidity.

There are many decisions to make regarding the right kind of sensor to use in which situation. For example, differences in manufacturing processes (e.g., process, discrete, batch process, etc.) have their own characteristics and sensed combinations that are unique to the production type. Other decisions about the type of device to use for sensing are driven by the requirements defined for the solutions. The array of sensing devices is as wide as the applications they serve.

Continuous process monitoring has different requirements than discrete processes. Combining data information feeds from a multitude of sources is often necessary to create a complete picture for performance assessment.

Key questions to ask are:

  • How do I configure the sensor to gather the right data?
  • How many readings do I take at what dimension?
  • What kind of environment will this sensor live in?

Why Should You Care About Source?

As the world of the IoT expands to address ever wider applications, its darker side becomes more exposed. Any time you connect an asset to an intranet via a sensor, you increase the plane of vulnerability of that asset. Therefore, sensors and devices require management, just as any other IT device. Managing these devices means that installation and configuration follow governance guidelines for provisioning, as defined by your enterprise. As your IoT implementation ramps to full deployment, challenges related to scale, security and connectivity can compound quickly, which, in turn, drive up costs and delay progress. Therefore, IoT deployment plans must include strategies for easy provisioning, configuration, monitoring, updating and decommissioning – full lifecycle planning applied to devices.

Because of their impact on the assets they are monitoring and their connection to factory control systems, source devices require security management as an active part of your secure operations guidelines. Some of the vulnerabilities to mitigate include unsigned firmware, default credentials, insecure data in transit, insecure key storage and missed firmware/software updates. Remember that pilots and proof of concepts are good learning exercises. Paying attention to key requirements here can save time and money later on.

CONNECT (Co)

METHODS OF EXCHANGING DATA

Of course, once you begin to access devices, it is necessary to set up connections to the enterprise network. A communications link or channel is a single transmission medium used to transport analog or digital information between a source and receiver. The main point of Connect is to build a communications architecture that enables everything from reliable machine operation to real-time analytics. Compare it to the nervous system of a machine, a plant, a company, and link the world. Commonly used transmission paths to support communications include guided media, such as copper wire and fiber optics, and unguided media, such as radio, microwave and lasers. A sound design involves the selection of communications technologies that deliver secure and reliable performance to enable freedom of source device selection and efficient application by machine experts.

Wired networks widely in use in factories today include PROFIBUS, PROFINET, Sercos, HART, CAN and Ethernet. However, running cable for wired networks is expensive and it’s difficult to reach some locations. Cable may need to be rerouted if the factory is reconfigured and it is not practical for equipment that is mobile.

Wireless technologies solve the issue of connecting mobile equipment and can be used to accommodate a flexible factory floor. They work well in most applications if the performance and features are aligned to the desired results.

Why Should You Care About Connectivity?

There are several challenges in wireless communications that need to be considered:

  • Reliability: meeting 99.999 percent reliability;
  • Robustness: operating in a radio frequency factory environment with extreme temperatures and vibration;
  • Determinism: accuracy of milliseconds given the current input and state when there is only one action that can be taken; 
  • Latency: addressing the millisecond delay between the controller and source device;
  • Security;
  • Durability in the factory setting in which it will be used;
  • Cellular plan charges: cellular and 5G requires paying service providers.

In addressing these challenges, several types of wireless applications are available that may make them more applicable to the usage cases you’re trying to solve. Zigbee Pro, Bluetooth®, Ethernet, cellular and LoRa® are all choices to consider, each offering unique characteristics to answer a wide variety of applications.

The introduction of wireless technology, whether in the smart grid or as part of the enterprise, control center, or remote site infrastructure, brings with it serious security concerns. Network health and cybersecurity are emerging failure modes that can be significant. Advances in wireless connectivity enhance flexibility and agility, but they also bring increased awareness of security protocols. The wireless network should be isolated from other networks by a firewall and implemented with strong encryption and authentication.

The U.S. Department of Homeland Security has documented the number of attacks mitigated by various security strategies. As a group of practices, they can be effective in preventing most intrusions.

Figure 4: Seven steps to effectively defend industrial control systems (Credit: National Cybersecurity and Communications Integration Center report, 3/24/18) 

COLLECT (Cl) 

PREPARATION AND STORAGE OF DATA

The main reason to collect data is to make it analysis-ready in context. Collecting and storing data is nothing new, but as you start looking at data as the fuel on which digitalized business models will run, collection takes on a whole new meaning. To perform diagnostics and hindsight analytics, mostly performed by human experts, the traditional historians and databases fed by time series data serve a purpose. The ever-increasing amounts and velocity of multivariate, multi-context, analysis ready data for digitalization requires new ways of processing, storing and preparing it in context in real time.

Why Should You Care About How You Collect Data?

There are more storage options and techniques being developed almost daily to handle the complexity, amount and dynamic nature of manufacturing data. Analysis of data is no longer limited to human analysis. Machine learning algorithms that may run in the Cloud or at the edge may need access to data collected anywhere else in the organization. Where should you collect and store the data? How do you maintain the original truth in the data as it moves around different usage applications? How will it impact the integrity of the digital twin that defines the lifecycle of the asset? What are the cost implications of retrieving this data, on demand, by different analytic applications? All these implications, and the increasing value of the data as an asset, require a much more careful and consistent approach to data collection.

ANALYZE (An) 

CONVERSION OF DATA INTO INSIGHTS

You seek insights from data so action can be taken. A natural question is: “I have been analyzing data to gain knowledge and insights before, so what is new about analyzing in an IoT world?” To date, most analyses have been performed by human experts. This limits an asset owner’s ability to analyze all the data, with the efficacy of the insight (i.e., conclusion) dependent on the level of expertise and experience of the expert. Additionally, the conclusion can be subjective. Most analyses have been focused on hindsight and diagnostics. Even when an analysis is performed for the purpose of foresight (i.e., prognostics), it is done as a discrete event.

In the context of digitalization, with more computing power and data analysis techniques, it is possible to analyze more data, create insights not possible before and create actionable prognostics and foresight. Algorithms can be trained to automatically perform analyses and the analyses can be continuous as new data streams arrive. This automation of analysis, performed by machine learning and AI, is a key differentiation to analyzing in the context of digitalization.

Why Should You Care About Analyzing IoT Data?

The amount of data generated by your assets has increased significantly in the last few decades. This amount of data will only increase further, exponentially, as new assets equipped with more sensors make their way into businesses. This data itself will become a valuable asset for any business. Rapid analysis of the data will make businesses more productive and profitable. How well you take advantage of the data through analysis and resulting action will impact the success of your business. For some businesses, the insights created by applying knowledge to data analysis can be monetized, resulting in new revenue streams.

DO (Do)

ACTION TAKEN FROM THE INSIGHTS

The main reason to analyze data is to make it actionable. You seek to do something that will return business value. Value derived from these actions can increase key business areas in productivity, revenue and quality. By observing the results of your actions, you learn and gain knowledge of best known methods and allow for continuous feedback improvements for your operations by extracting the most value from your digital and physical assets.

Why Should You Care About Do?

Industry today spends a vast amount of money on IT infrastructure, like networks, data storage and applications. Companies need to ensure that the expense is delivering value by returning actionable and measurable results. Willing or not, organizations have been thrust into the digital age and, as a by-product, have captured unwieldy amounts of data and information. With this, analytics has become a necessity to glean insights from the mountains of data. However, insights from analytics in and of themselves produce value only when one can do something valuable with them.

Figure 5: Rapid analysis of data will make businesses more productive and profitable

Turning these insights into action can be achieved with promising advances in technology, like AI, including machine and deep learning; autonomous operations; human augmentation; and electronic workflows. These technologies allow for analytics to be put into action to measure the resulting effect and self-learn. The analytics to action feedback loop is validated when the do action produces desired results. You then learn from your data and can then extract the most value from your assets, both physical and digital.

The Digital Culture of Your Organization

For digital transformation, people need to embrace a culture of digitalization, where digital processing is preferred over manual. Leadership is key in driving digital transformation to meet business goals; sharing the vision as a competitive differentiator and seeding cultural changes to shift toward a digital culture.

Are You Ready?

If organizations are to advance from the recognition phase of the potential of AI and the IIoT to taking action, improvement of foundational elements of all five Uptime Elements knowledge domains is required. All domains have an impact, but especially:

  • Asset Management;
  • Strategy and Plans;
  • Decision-Making;
  • Asset Lifecycle Management;
  • Risk;
  • Asset Knowledge;
  • Cross-Functional Leadership for IT/OT/ET.

Creating a holistic approach to advancing reliability and asset management built on a foundation of reliability leadership culture across information technology (IT), operational technology (OT) and enterprise technology (ET) is needed to realize the full promise of digitalization. Guiding the vision of the IoT with the digitalization strategy framework in mind will create credibility at the same time.

Realizing the value of data is priceless. As you create a transparent, decision-making framework and enhanced asset knowledge, you’ll generate new insights cross-functionally and often outside your own organization.

To prepare for the future of work:

  • Invest in new competencies and processes as you enhance human capital management, with a focus on enabling and empowering humans to work with digital information for machine-assisted reliability and asset management. Use the Uptime® Elements — A Reliability Framework and Asset Management System™ as a guide.
  • Help to shape the workforce of the future. People will always come first before technology. Use your reliability leadership as a foundation and suggest/implement organizational changes that support a cross-functional and collaborative working environment enhanced by AI, machine learning, augmented reality and the IoT.

If you do not have a destination, any road will get you there. Create a policy, a strategy and a plan that includes digitalization and create a line of sight to the aim or objectives of the organization.

What’s Next?

For the full story, attend the Maintenance 4.0 Forum, co-located with The RELIABILITY Conference™, from May 6 to 10, 2019, in Seattle, Washington. The authors of this article will hold a complete workshop on the Uptime Elements IoT Knowledge Domain and Digitalization Strategy Framework. For more information visit www.reliabilityconference.com.

From Your Site Articles