Data is everywhere these days and I see more and more people taking data seriously when it comes to the business and organization of engineering systems. Back two decades ago, CAD file was a foundation of engineering data. In my life, it was AutoCAD, which became the first example of an engineering platform. AutoCAD files were ultimate data storage for everything construction and manufacturing companies needed. Additional applications stored data in AutoCAD files and it eventually grew as a place to collect all engineering information. AutoCAD was a powerful system. Similar things happened with other CAD systems. I’ve seen similar processes happened with other CAD systems – Catia, Pro-E, and later Solidworks and others. The nature of design and engineering processes made data located in CAD systems naturally.
The ultimate data architecture of the engineering system back in the 1990s was presented as three layers – (1) File Management; (2) Geometry and (3) Metadata. The (1) and (2) were connected inside CAD files that required management – storage, versions, aggregation, and sharing. The communication was heavily dependent on exchanging of documents and CAD files. Although the future development of PDM (and PLM) systems was always focused on how to get data out of CAD systems and made it available downstream for other departments and organizations, it was always very much focused on the documents. The CAD files and their derivatives were fundamental parts of all these systems.
File-Based Platform Challenges
File-based approach and data management systems focused on files and their derivatives were growing up naturally. But they have some fundamental flaws. The main problem of the approach is to maintain file type compatibly and data consistency. Eventually, companies need to work with many systems and each of them presents its own needs to the files and might need to save different information. Which immediately makes the file-based approach questionable. There are many examples of inefficiency related to file-based systems.
Databases and Data Management Platforms
Problems of files based approach made companies to re-think their system architecture in data management. To get rid of files and turn them into the database approach is not a new idea. In a variety of forms, the idea was presented in the work of several CAD and PLM companies.
Dassault Systemes spent a substantial amount of resources and time moving CATIA V5 to a file-less data-driven platform. Initially called CATIA V6, it made its transition to what is known today as the 3DEXPERIENCE platform (3DX). The new 3DX platform is managing data (not files) and introduces a way to store objects in the database. Clearly, it provides a better way to manage information and relationships. The following InFlow article is a simple explanation with no much marketing about the difference between CATIA V5 and V6 approaches. The magic word “data-driven” is used at DS as a keyword to differentiate from previous V5 file-based architecture.
A major advantage of having a Data-Driven platform is that it gives a “single version of the truth”. This means that modified data will be updated across the entire platform in real-time. The data is ALWAYS up-to-date. For example, if two designers are modifying different parts within an assembly at the same time, each designer can see the modifications that the other has made without having to close, re-sync, and re-open the assembly. The user needs to simply perform an “Update” command. This also means that whatever data the supplier or manufacturer can access is always the latest. This circumvents the issue of wasted time by having worked on outdated data.
Another example of transformation into placing the “Data First” approach is Autodesk. As a company with many product lines, Autodesk is suffering from multiple files and transitions needed to perform even inner product workflows. For the last decade, Autodesk is transitioning to the cloud platform (known as Autodesk Forge) and the elimination of files is part of the goals.
Here is one that I found interesting. AEC Magazine brings examples of Autodesk developing an online platform for construction combining information from architecture and construction systems such as Revit with information from manufacturing systems (eg. Autodesk Inventor). Check the article here. And this is my favorite passage
Autodesk is creating a new cloud platform for Industrialised Construction, which will be a collaborative ecosystem where Autodesk’s product portfolio and products from other vendors will be able to integrate. While the demos given concern Inventor and Revit, the reality is many firms in fabrication are using products like Solidworks to model their parts. However, Autodesk is in charge of Inventor and Revit development and can work towards them playing better together, in this case through a technology layer in the cloud.
….Marks gives visual examples of Autodesk’s unreleased ‘Integrated Construction’ platform. This is particularly fascinating as it shows how the industrialised construction web service, takes in an Autodesk Revit model, allows the production engineer to pick the parts to be manufactured (termed Scope) from the Revit model viewer, enables easy configuration of those components through templates and automates the creation of fabrication-ready Inventor model, drawings and Bill of Materials (BoM) (this was a feature in Inventor 2021).
An interesting aspects of the platform is to get rid of drawings and replace it with product-oriented data. It is not clear although how the project will be dealing with the data – will it create a neutral data representation or will continue to transform files between different environment.
A similar problem described above exists in many other engineering applications. The demand to provide a better approach to managing information in a more granular way, to make information access streamlined between departments and organizations, combined with the opportunity to provide intelligence based on the data, brings the question about future engineering software architecture for both construction and manufacturing companies.
The process of transformation from a file-based approach to a data-oriented approach is happening everywhere these days and cloud-based systems are playing the role of the catalyst in this transformation.
I can see three main elements of the transformation from file-based architectures to engineering data platforms.
1- Granular Engineering Data Model
The data in design and engineering systems is super diverse. Product information is a complex beast and making a system capable to handle the complexity of the data is not a simple task. Manage files was easier, but when it comes to the details, granularity is the key. The data is CAD systems and related systems need to manage, co-exist, intertwined, and managed from a variety of perspectives (eg. complex configurations, change management, etc.)
2- Multi-tenant architecture
CAD files are easy to transfer between users and organizations. New platforms should be able to do the same without losing the data consistency and, at the same time, not require to put organizational boundaries. So, multi-tenancy is a key element. Without that, files will be a transport vehicle and all problems I described above in file-based systems will stay.
3- Flexible and expanded user experience
Simplicity is hard. Complexity is a killer of most CAD, PDM, and PLM systems. Don’t get me wrong, engineers like complexity. You won’t buy engineering love with the simplified systems not capable to manage the complexity of the information. At the same time, user experience is extremely important, and a new generation of users like systems to be more intuitive. Combined with the high diversity of the data, it creates a big need for flexible and expandable user interfaces.
What is my conclusion?
Data is winning. The 2020s will be a decade in which engineering platforms will be moving from file-based approaches to granular engineering data platforms. Life without files would be much easy. Watching my kids considering Google Docs as a perfect abstraction, I understand that. But files are not going away any time soon. The legacy will be dying slowly together with people that were born to use File-Open/Close/Save commands. Just my thoughts…
Disclaimer: I’m co-founder and CEO of OpenBOM developing a digital network-based platform that manages product data and connects manufacturers and their supply chain networks. My opinion can be unintentionally biased.