Every industrial business—indeed, every conceivable kind of organization—runs on data. Or to be more precise, they run on the insights generated from the enterprise’s data, making key decisions based on those insights and their value in the appropriate context.
Yet everyone seems to agree that most of what is in our databases is of questionable value, no matter who owns the data or what it’s used for.
So, here’s the “running-on-data” problem. On the one hand, data is essential in running any business. On the other hand, much of our data may be worthless or inaccessible to those who need it. How can these two apparently incongruent statements be reconciled?
I believe the answer lies first in placing each discrete bucket of data in context, such as where it was created and by whom (i.e., person and/or system) and where and by whom it can be accessed. Decision-makers can then extract value from the data and the depth of insights needed for sound decisions. Even with this, uncertainties will persist.
How do we know? How can we be sure? These questions should lead us to dive into our data periodically to ensure its context and associated accuracy. As enterprises transform themselves digitally, they gradually shift away from insights based on direct, hands-on experience. As we rely more and more on insights readily extracted from data, periodic fact-finding efforts become increasingly important. Otherwise, our decisions may be less than optimal.
You do not have to read them in any specific order, but this is one of four self-contained articles written by CIMdata for engineering.com. The initial article was entitled “Top 12 Trends and Enablers to Unlock Every Digital Transformation.” The next article released covered how, and why, Digital Threads in PLM are really webs of data. The remaining article will explain what digital transformation requires of the workforce.
Digital Transformation Reality Check
Amid your enterprise’s long march to digital transformation, are you managing your intellectual assets like any other critical business asset? Is anyone held accountable for database content and data accuracy? Do any repeatable data governance policies and procedures exist? If so, are they followed? Digital assets, in contrast to physical assets, cannot be touched or handled and so are often poorly managed.
I wish to address the digital transformation challenges in five of CIMdata’s Critical Dozen. For reference, these evolving trends and key enablers of digital transformation are:
- End-to-end connectivity
- Data and process management
- Configuration management (CM)
- Bills of information (BoI)
- Model-based structures
- Digital thread
- Digital twin
- IoT and product lifecycle management (PLM)
- Changing views of the “product”
- Big data & analytics
- Data governance (DG)
- Digital skills transformation
Each addresses the paradox at the heart of this article. Yes, unfortunately we are often dependent on questionable data, but digital transformation lets us do something about that.
Digital Transformation Reality Check: Data and Process Management
Start with the need to enable consistent data and process management approaches and supporting technologies to ensure that the information held in business-unit repositories is available and always ready for use.
As we know, many of these supporting technologies are legacy software packages and departmental “silos” that hoard and lock away task-specific information. Outdated practices, such as manual data re-entry and tedious workarounds, are perpetuated by silos and legacy solutions. Eliminating them helps to unblock the bidirectional data flows required by the enterprise’s linked, enabled and cross-connected processes.
Modernizing data and process management also gets rid of most of the difficulties associated with access, reuse and analysis of critical information.
Digital Transformation Reality Check: Model-based Structures
Implementing model-based structures is the next critical element. As a central part of digital transformation, model-based structures are the new data repositories, ending the use of text documents, 2D drawing files and the constraints imposed by their formatting.
In the model-based enterprise (MBE) that emerges, multi-dimensional structures—functional models—represent every component, tool, system, product, service, function and process used by the enterprise. Model-based structures support the countless moving parts of digital transformation by greatly speeding up innovation and data management.
As model-based structures come into everyday use in product definition, design, product engineering, systems engineering, manufacturing and even service, productivity will grow and errors will dwindle.
Digital Transformation Reality Check: IoT and PLM
The inherent connection of the Internet of Things (IoT) and Product Lifecycle Management (PLM) comprise the third element in understanding and enabling a true and long-lasting digital transformation.
The IoT’s endless flows of data are essential to 21st century innovation. These data flows, which are potentially overwhelming, are best handled with a robust and comprehensive PLM strategy—ideally in an enterprise-wide systems-of-systems architecture. When properly linked to PLM’s digital threads and digital twins, the IoT’s connected devices, software and sensors offer nonstop insights about the performance of products in customers’ hands.
With the IoT and its more focused subset, the Industrial Internet of Things (IIoT), PLM can help designers, managers, and top-level decision-makers avoid muttering in frustration by closing the loop between the physical and virtual world of a company’s products and services.
Digital Transformation Reality Check: Big Data and Analytics
Big Data and analytics deliver the much-needed capability to show us what we have been missing when we look at our data. They are a crucial fourth element in finding the big picture hidden in our data, often unearthing nearly undetectable unknowns.
The mass and variety of Big Data that is available to us is the cause of the struggle of today’s widely used analytic tools and methods that make sense of it. The volume, complexity and continuous change of Big Data overwhelms users—unless they have robust and scalable analytics solutions.
Serious consideration should be given to topological data analysis (TDA) and other AI-enabled Big Data approaches and tools. TDA is impressive at finding hidden patterns and relationships in unstructured data, helping to uncover “unknown unknowns,” as well as the less-challenging “known” unknowns that we may be aware of.
Digital Transformation Reality Check: Data Governance
The next element in unearthing the Big Picture in our data is data governance (DG). It ensures that the digital transformation gains of the preceding four topics don’t unravel.
DG provides end-to-end oversight and sustainability of essential information, including transparency, accountability, ownership, sound day-to-day management policies and effective long-term stewardship. These all help coordinate and reduce friction in change management, product development, analytics, support and much more.
How to See the Big Picture of Digital Transformations
The diametrically opposed axioms at the beginning of this article highlight the reality that the data processing revolution—digital transformation—has proven to be a mixed blessing.
We are blessed with the capability of quickly understanding large amounts of information and making productive, insightful use of it. But comprehending what is happening around us in the broadest terms seems to go missing if we rely exclusively on data for insights. Examples include unanticipated changes in customer preferences, competition, new markets, service requirements and resource availability that, with hindsight, were foreseeable.
Decision-makers addressing the big picture must guard against any urge to ignore the experience and the intuition of seasoned workers and managers. If we allow databases to make it easy for us to sidetrack experience and intuition, the resulting loss of comprehension will be much more than collateral damage.
In other words, we must not take people out of the equation. They create information; they do not simply use it. More importantly, they convert information into actionable knowledge.
The elements I describe here and in my other writings are not panaceas. But with focused effort, these elements are critical components of a long-lasting and successful digital transformation.
Design, development and sustaining innovative products and services requires understanding what information is needed by the developers of each product and how to retrieve it from the enterprise’s oceans of bits and bytes. These requirements are the same for all individuals and systems that operate throughout the entire lifecycle—from concept through end of life.
At the same time, we need to be sure we can uncover and comprehend the big picture. In business and industry, what we see as understanding our data may be short-sighted. Looking three to five years ahead to anticipate products, services, competition and resources in the market requires comprehension.
Bearing in mind that in war, truth is the first casualty (Aeschylus, 525 - 456 BC), we must not let intuition and hands-on experience be a casualty of data processing.
So, now is the time to sort out, clean up and—as resources allow—identify and delete useless data. Start with the oldest files. If in doubt, ask for help. Don’t be shy about asking for funds and expertise. Make sure all new data is checked, sorted, archived and regularly updated to block any dubious input.
Then correctly apply the critical focuses on data outlined here.