What is the next big topic in PLM technology, PLM systems, product data management and product development process modeling? Speaking to many PLM professionals, I can hear the word “graphs” is coming more and more often in the context of modern PLM solutions and modeling of business processes. Product’s lifecycle is getting more complex and it demands a better tech to support the development of modern PLM software. What is behind “usage of graphs” and how it can impact future development of PLM systems? How it can transform PLM software, architecture and technologies? Can it provide a better user experience? Let’s talk about it.
Why Graphs are Becoming Popular?
Graph technology is gaining widespread recognition in technology and industry. As products become more complex and connected, the relationships between data points—components, processes, suppliers, and more—become harder to manage with traditional abstraction models. Graph technology offers a flexible and powerful solution by structuring data as nodes and relationships, which mirrors the complexity of modern products. This shift has led to an increasing number of companies exploring how graphs can enhance PLM processes. However, the devil is in the details. What means “let’s use graphs in PLM software” and how it can make an impact.
Previous Discussions on Graphs and PLM
I published quite a few articles related to graph technologies, semantic web and PLM in the past. In these articles, I’ve discussed the growing importance of graphs in PLM and how they provide a more flexible and scalable solution compared to traditional hierarchical databases. Graph technology allows you to manage and analyze the relationships between various elements more effectively, improving collaboration, design management, and decision-making across the lifecycle of a product. Here are a few links you can check out.
Social PLM, Graphs and Organizational Overlap
Why graph analyzes will rule PLM in the future?
Graph browser and PLM user experience
Graph Mania – 3 Options To Include Graphs and Networks in PLM Architectures
From Digital Thread to Holistic PLM Graph Models
The Importance of Knowledge Graph For Future PLM Platforms
Graph Databases, GQL Standard, and Future of PLM Data Layers
I wrote them in a different time for the last 10-15 years. You can google more articles I wrote about graphs, semantic networks, and their applications in product lifecycle management technologies.
Not all graphs are the same – usage of graph technologies in PLM
Despite the growing popularity of graph technology, there is still a lot of confusion in the industry about what “graphs” actually mean in the context of PLM. Different vendors and platforms use the term in various ways, which can make it difficult for companies to understand what level of graph support they actually need.
In this blog, I’ll breaking down graph usage in PLM into the following three main maturity levels:
- Graph-based user experience.
- Graph models and abstraction layers.
- Graph database technologies.
By understanding these levels, you can gain a clearer picture of how graphs can impact your PLM system and its capabilities.
1. Graphs and User Experience
The first level of graph technology in PLM is how it improves user experience (UX). PLM systems with graph-based interfaces allow users to easily navigate complex relationships between parts, assemblies, and suppliers. Visualizing data as a network of interconnected nodes makes it simpler to find relevant information and identify patterns or dependencies within product structures.
For example, with a graph-based UI, engineers can quickly trace how a change to a single component impacts multiple assemblies or products. It enables more intuitive interaction with the data, reducing friction in decision-making and collaboration across teams.
2. Graph Models and Abstraction Levels
The second level is the use of graph models and abstraction layers to describe product information. This level moves beyond just the user interface and into how the system models relationships between data.
Flexible data modeling was an important layer in building PLM systems for many years. While PLM traditionally were using RDBMS, the object layer and relationships were flexible and I can see a growing usage of graph-based modeling. With graph-based modeling, you can model product information more naturally by using flexible, interconnected networks of nodes and relationships. This abstraction layer allows you to:
- Model complex product assemblies and their dependencies.
- Seamlessly connect product data with manufacturing, supply chain, and service information.
- Build more adaptable models that can evolve as product designs change.
This level of graph technology provides the foundation for a more dynamic PLM system, where changes are reflected across all related components and information sets in real time. However, it is still an object model. To have a full leverage of graph technology, we need to move to the various graph databases and query languages.
3. Graph Data Modeling and Graph Database Architecture
The third and most advanced level involves graph databases and how data is stored, queried, and managed. Graph databases are purpose-built to handle vast networks of interconnected data. Unlike relational databases, which use tables and fixed schemas, graph databases use nodes to represent entities and edges to represent relationships. Graph databases has grown their presence for the last 10-15 years. Here you can see the list – Graph Databases (Wikipedia)
From my perspective, graph database provide a very powerful database storage for interconnected datasets used in PLM systems. Here are three reasons why graph can be an element of modern PLM architecture.
- Efficient management of product structure (eg. BOM, design structures, etc.)
- Fast querying of complex relationships across products, suppliers, and processes.
- Scalable solutions that can handle the growing complexity of modern manufacturing data.
Several types of graph databases exist, such as labeled property graphs and semantic graphs (so called RDF graph). These are more often used to model knowledge and ontology. Each type of graph database comes with its own set of strengths, depending on the use case. The databases for different types of graphs can be different, but the common theme here is that with graph databases, PLM technologies are getting a native advantages of graphs – fast queries, graph data science, usage ontologies and other tech that can make PLM system much faster and more expressive.
What’s Beyond? Graph Data Science and AI Applications
Graph technology doesn’t stop at just data storage and modeling. The next frontier is graph data science and AI applications. By leveraging graph technology, PLM systems can integrate machine learning and AI capabilities to analyze data in ways that were previously unimaginable.
For instance, graph-based AI can be used to:
- Predict product failures by analyzing historical data and relationships between components.
- Optimize supply chain networks by understanding how changes in one supplier affect others.
- Enhance decision-making through automated reasoning over complex product data relationships.
I’ll be covering these advanced applications of graph technology in PLM in a future blog.
What is my conclusion?
It is important to understand details about usage of graph technology in PLM. Graph technology is becoming an essential tool in the world of PLM, but it’s important to understand what different companies and vendors mean when they talk about “PLM graphs.” Whether it’s improving user experience, modeling complex relationships, or leveraging graph databases, each level of graph support offers distinct advantages. As this technology continues to evolve, understanding these levels will help you make informed decisions about adopting graph technology in your PLM system.
Just my thoughts…
Best, Oleg
Disclaimer: I’m the co-founder and CEO of OpenBOM, a digital-thread platform providing cloud-native PDM, PLM, and ERP capabilities. With extensive experience in federated CAD-PDM and PLM architecture, I’m advocates for agile, open product models and cloud technologies in manufacturing. My opinion can be unintentionally biased.
