Engineering and manufacturing systems are moving to the cloud. Born in the cloud, “cloud-native” systems are more efficient, scalable, and reliable than traditional systems. A big part of this shift is data-centric architecture. Data-centric architectures move away from monolithic data management systems towards granular and flexible data model that allows more granular access to the data and operations. This allows for better scaling and performance when data grows or changes over time. Cloud-native systems and data-centric architectures are key to the future of manufacturing. In this blog post, we’ll discuss what they are and how they can help your business succeed.
Cloud technologies changed the landscape of many industries for the last 10-15 years. I think it is actually really hard to find any new software these days that is not cloud-based in some way. The software is connected, operates from browser and mobile devices, is updated automatically, and provides a variety of functionalities that support communication and data sharing between users and organizations.
However, in the engineering world, the situation is very special. The software lifecycle in CAD and PLM software is extremely long and solutions are enormously complex when it comes to the selection process and criteria. CAD and PLM is probably the last software industry where debates about the usage of cloud technologies are still very hot and cloud marketing is in the air these days. Companies are using cloud marketing tools to convenience the advantages of their solutions.
In my article today, I want to talk about, cloud platforms, cloud-native architectures, cloud strategy, CAD, and PLM application development and compare how these systems compare to old-fashion monolithic systems. Digital transformation makes cloud-native systems an important element of future cloud adoption and raises the question about future CAD and PLM software development.
Cloud-Native Application
My attention was caught by Onshape’s article – Don’t be fooled: Cloud Native vs Cloud Storage CAD solution, which speaks about characteristics of software solution and architecture that makes a specific solution “cloud-native” as well as some ‘warning signs that you’re on a cloud-storage-based solution.
In a cloud-native platform, all your data is stored in one place in a database in the cloud. When a user works on data, it does not need files to be copied, locked, or downloaded. Your data is also stored incrementally, which means any state of data that ever existed can be easily restored.
Your software, a collection of microservices, also runs entirely in the cloud. Microservices are multi-tenant, allowing rapid scaling and instant replacement of “broken” microservices. This is a fundamentally different approach than a cloud-storage approach.
I liked the 25 signs that mostly outline Onshape specific features and functions such as an ability to manage data in a centralized form and provide granular access to the data for different users. Onshape architecture is indeed unique in the market. While 25 use cases are super descriptive to explain what can go wrong with any system that is not similar to Onshape, I found these 3 things summarizing Onshape architecture.
1. No files, data is managed “as a whole” in the database
2. Simultaneous data access by users based on their access control
3. Full history of operations is preserved and allowed data to be restored
What is important in the architecture offered by Onshape is these three data principles. Combined with browser-only access, they represent what Onshape marketing calls “cloud-native”. The elements of this architecture represent a very interesting evolution of CAD and PLM architecture that can help to describe what the future of PLM architecture will look like.
Data-Centric Collaborative Lifecycle Architecture
All cloud-based applications are generally using databases to store the data. So, from that standpoint, you can think about all “cloud” or SaaS systems very similar. However, here is the thing… Onshape replaces files with a very specific mechanism allowing changes managed in a very granular way and these operations are recorded in the databases. The so-called “workspace” iis the mechanism that allows an Onshape application to collect changes and to perform collaboration between users. The architecture is probably unique in CAD applications, but what is interesting is how other applications (both in PLM and outside of the PLM domain) might be using the same architecture to redefine the architecture and functions of many systems.
The key element of new cloud architecture is redefined data management layer with the following three important characteristics:
1- Data management layer with granular operations
2- Simultaneous multi-user access
3- Special type of lifecycle to support the above two mechanisms
A special type of lifecycle includes the ability to capture data in the logical business operations that must be supported by the system. Here are a few examples of similar systems that work like Onshape outside of the CAD world. Atlassian JIRA (https://www.atlassian.com/software/jira), one of the applications mentioned in the Onshape article will automatically preserve all changes to the ticket instantaneously (including comments) and will allow users to use them later even if the operation was interrupted. OpenBOM (openbom.com), allows to preserve the data in “latest state” in Item and BOM and provides a mechanism to save it as immutable revision. GitHub (https://github.com/) “Trunk model” allows merging all changes made in a source code in the centralized storage including all bug fixes, features, and other changes.
Micro-Transactions And Collaborations
An interesting observation of all architectures I discussed above is related to two main elements of the technologies used in all these systems: (1) granular data transactions; (2) collaborative application and user experience. The last one is extremely important. In many existing CAD (and PLM) applications that collaborativeness is generally limited by transactions of data management foundation of the system. Such a “file-based” CAD system can only preserve the current state of the file, but cannot go to the part of the file (not enough granularity). The same about collaborative PLM applications limited by lifecycle operations in PLM databases.
What is my conclusion?
Cloud-native applications will play a key role in the future of CAD/PLM production development. A new class of data-centric architectures and data management and collaborative applications is coming to CAD and PLM space. The main enabler of these new applications is a rethinking of how these applications are managing data and support lifecycle (change) processes. It introduces a new set of cloud-friendly functions, eliminates the need to manage files and to be limited to old fashion database transaction mechanisms, and allows support for new ways to collaborate. The note for all PLM architects is to closely watch technologies and be careful with flashy marketing messages about cloud and digital platforms. Just my thoughts…
Best, Oleg
Disclaimer: I’m co-founder and CEO of OpenBOM developing a digital network-based platform that manages product data and connects manufacturers, construction companies, and their supply chain networks. My opinion can be unintentionally biased.
