Advanced robot arm system for digital industry and factory robotic technology . Automation manufacturing robot controlled by industry engineering using IOT software connected to internet network .
Digital transformation has unfurled a new chapter for the world of manufacturing and product lifecycle management (PLM). This seismic shift, predominantly powered by data, brings both promising avenues and daunting challenges. Digital transformation changes the way we use technology for product development and manufacturing processes. And this process put data in the middle.
The days when manufacturing operations were solely reliant on manual processes orchestration and rudimentary raw materials data management technologies like spreadsheets (or cork boards) computer systems are slowly, but disappearing . Today, the wave of digital transformation is sweeping across industries, profoundly altering the landscape of product development and manufacturing.
Central to this transformation is data. Data has emerged as a foundation in the manufacturing transformation. Started from engineering BOM and manufacturing BOM management, later expanded with more holistic material requirements planning (MRP) and product lifecycle management approach (PLM), companies are able to control their processes, calculate materials shortages and control BOM structure of products.
However, while the potential of data is undeniable, harnessing its power is not without its challenges and doesn’t stop with engineering bill and manufacturing bill. The opportunity of smart data usage goes much beyond than just managing of an accurate BOM with raw materials and various components and materials required.
In my blog today I want to talk about why collecting data is not simple, what to do about it and who is ready for a data-driven world?
Why Not All Data is Born the Same?
Contrary to popular belief, data collection isn’t a straightforward task. My attention was caught by Jan Bosch blog – There’s no such thing as “the data” speaking about complexity of data and importance of the context in data collection. I like examples Jan Bosch brings to explain that collecting data is a complex process, which can lead to unknown results. It’s essential to recognize that not all data is created equal. The sheer volume, variability, and velocity of data make its collection and interpretation an intricate endeavor. Here is the passage:
There’s no such thing as “the data.” Instead, each question we seek to answer and each variable we look to track over time needs specific data to be collected from systems in the field. In practice, historical data often lacks context and is highly limited in volume. In addition, confounding variables complicate analysis to a significant extent.
So, although companies are literally drowning in the data, it might need to do some “digging” to bring this data in the form when it can be used for the purpose of optimization of product development, manufacturing and other business processes.
Data Can Help, But How to Make it Happen
Given the complexities surrounding data collection, how then can organizations leverage this asset effectively? Brining data in the central location in the cloud seems to be a solution. For many years, a “single source of truth” mantra was a main PLM (and other enterprise systems in manufacturing) selling point convincing companies that by brining the data in a single database, it will allow to use the data and streamline processes.
However, the reality of PLM developed for the last decades convinced companies that it is impossible to bring all data in the same system. At the same time, the value of the data and, especially, recombining data from multiple sources can be huge. Industry Week published the article by Andrew Anagnost, Autodesk CEO – Masterful Use of Data Is Manufacturing’s Future: Who’s Ready? The article brings an interesting point about how open and accessible data available on demand can change manufacturing similar to how navigation systems changed the driving process for the last 30 years.
Here is an interesting passage from Andrew Anagnost artice.
Recently I met with a German manufacturer who told me his company is facing $20 million in unbudgeted costs next year. The world is less certain than it used to be, and while unpredictable conditions demand dynamic responses, many manufacturers overlook a valuable tool to help them quickly make informed decisions: accessible data.
Open, accessible data—on-demand, centrally secured in the cloud, easily connectable and extensible—can help solve manufacturers’ biggest challenges. The more data in a system, the better it performs. And when artificial intelligence pulls from anonymized, aggregated data shared by multiple sources, it can make an entire ecosystem more effective.
Think of how driving in an unfamiliar city has changed thanks to navigation apps, which rely on anonymized data processed by AI. Traffic apps help everyone use known, existing infrastructure more effectively and avoid compounding bottlenecks. They do so, however, without disclosing private information like why someone is moving from one place to another, what they’re driving, or what they’re carrying. But each driver using the system is kept up-to-date on changes as they occur in real time – a potentially huge time-saver.
Similarly, a new generation of cloud-powered software allows manufacturers to create and store their design, simulation and manufacturing knowledge in a single place. Overlaying datasets from multiple sources can help more dependably track fluctuations in the costs of materials, energy, shipping, manufacturing methods, labor and more. These forecasts can guide decision-making about choosing renewable materials, for instance, or different suppliers—and whether a product can be manufactured more simply, or closer to where it’ll ultimately be used.
My key takeaway from the article and the main element of Autodesk strategy I learned recently is to unlock data from closed formats and lift people up from silos, which will lead to the capabilities of collecting, centralizing, and analyzing data from its many systems and business processes, ranging from designs in CAD, PDM, PLM, ERP, CRM, risk assessment and quality management tools.
However, to make it happen require new type of systems and data management architecture. Different from existing PLM legacy that was developed 20+ years ago.
Is PLM Industry Ready For Digital Data Transformation?
Can PLM software vendors lead the way of providing data management solutions collecting, aggregating, normalizing and organizing data that can be used for business decision? When we look in more details on the PLM software industry, a contrasting picture emerges. While there’s a huge buzz around cloud solutions, the PLM industry largely remains focusing to hosting existing software on cloud platforms like AWS, Azure, and GCP. Therefore the question comes – can merely hosting PLM software on the cloud suffice in harnessing the transformative potential of data-driven manufacturing?
In my view the PLM industry and vendors stand at a crossroads. While the allure of the cloud, modern data management, data analytics and AI is undeniable, the reality of enterprise PLM platforms leans more towards merely relocating existing systems to platforms like AWS, Azure, and GCP. You can check 3D EXPERIENCE, Aras Enterprise SaaS, Teamcenter X, Windchill+. All these systems are hosted versions of their “on premise” PLM siblings. This raises a critical concern: Is current PLM strategy a merely a cosmetic upgrade? Does shifting traditional PLM software to the cloud genuinely harness the digital transformation’s potential, or is it just old wine in a new bottle? How to bring new data management architectures to enable the future Andrew Anagnost’s article is speaking about?
How to Organize and Collect Data In The Right Way?
The key to unlocking the potential of data in manufacturing lies in its effective organization. Organizations must emphasize creating comprehensive information models to contextualize data, constructing knowledge graphs that foster meaningful relationships. This approach can bridge the data silos, offering a more holistic understanding of processes and catalyzing data-driven insights.
Such methods will allow to bring data in the context, navigate between silos and support processes by providing the right data at the right time. To truly tap into data’s transformative power, a modern data management approach to its organization is paramount. This involves:
- Rich Data Models and Contextualizing: Building flexible data models that align data with real-world applications and scenarios.
- Building Knowledge Graphs: By mapping data interrelations, knowledge graphs can unearth patterns and insights that isolated data points can’t. KG is foundation for future AI systems.
- Fostering Interoperability: Data must flow seamlessly across departments and systems. This involves integrating different data silos, ensuring a comprehensive data view.
Modern Generation of PLM systems
Addressing the aforementioned challenges necessitates a radical re-imagination of PLM systems. What’s needed is a new generation of data management platforms and PLM software, one that prioritizes contextual data modeling and knowledge graph construction. Such systems can harness data’s potential by facilitating its collection in context and fostering semantic relationships, thus driving richer insights.
To make it happen PLM architectures must go through the generational shift. The future PLM systems architecture must:
- Prioritize Flexible Data Modeling: To understand data, one must understand its origin and application. Contextual modeling does precisely that.
- Incorporate Knowledge Graphs: Instead of rigid data structures, flexible knowledge graphs can adapt to evolving data landscapes, capturing nuances and fostering deep insights.
- Adopt Multi-tenant Data Architectures: All legacy PLM enterprise systems are single tenant, which provides significant limitations to managing data in a large and complex systems with OEMs, suppliers and contractors. Such multi-tenant architectures will allow to isolate tenant data and, at the same time apply AI and other technologies to recombine data from multiple silos and locations.
What is My Conclusion?
Manufacturing industry is at the beginning of the process of understanding how to build a data-centered manufacturing process. There is an acceptance of the cloud-based software and early understanding of processes that can be established using data. However, there is a gap between old-generation of PLM software and the potential of data-driven processes. PLM software and PLM vendors must close the gap to make PLM software platforms to stand for the demand of manufacturing industry for transformation and future use of digital data. Just my thoughts…
Best, Oleg
Disclaimer: I’m co-founder and CEO of OpenBOM developing a digital thread platform including PDM/PLM and ERP capabilities that manages product data and connects manufacturers, construction companies, and their supply chain networks. My opinion can be unintentionally biased.
