With Cloud technologies developing rapidly in recent years, so came the thought of using the Cloud for hosting various systems in it, from computationally demanding, through mission critical and more. Product Lifecycle Management (PLM) systems were no different and a large number of both PLM vendors and customers attempted to leverage benefits the Cloud brings to gain that competitive edge world’s best manufacturers always strive for. Does this approach make sense and what benefits (and drawbacks) does it have over traditional on-premise hosting? What are the differences between various XaaS (“Something”-as-a-Service) models for Cloud PLM solutions? How can we make sure that a Cloud PLM system is efficient both cost- and performance-wise?
Introduction to Product Lifecycle Management (PLM)
Product Lifecycle Management (PLM) systems are created to control the life cycle of a product and all data connected with – from idea to implementation, and even beyond. Nowadays, these systems are a key tool in most product design and manufacturing companies. Featuring BOM management, change management, collaboration tools, workflows and even compliance tracking and supply chain management, PLM significantly improved the information flow within organizations and facilitated collaboration, even between teams dispersed around the world. This in turn allowed shortening the time of introducing new and updating existing products.
PLM is the foundation of any modern design process. Although as a concept PLM is very efficient, PLM systems were historically quite complex and demanding (from a resources perspective). Suddenly thousands of engineers began feeding fresh CAD files into web systems, which triggered reviews, approvals, changes, publication, etc. generating massive amounts of data in the process.
As you can imagine, infrastructure requirements for most advanced PLM systems were quite hefty. Not only large amounts of storage were needed (for databases, file vaults), but also quite a few machines for the application itself (sometimes requiring several nodes to handle large loads), publishing engines (so-called “workers” for CAD and other document types), and so on. Whenever you need additional servers (for example: development environments), scalability issues surface, as you often need to either plan resources ahead or acquire them ad hoc. Add to that the need to back it all up once in a while (you are doing backups, right?) and it turns out you need a pretty expensive infrastructure to run it and a pretty large IT department to manage it all.
Theoretically you could tackle at least the last requirement (IT) by automating some stuff, but unfortunately, it’s not that easy when you are dealing with physical infrastructure with perhaps only basic virtualization.
“So,” you may ask, “is there a way to make this thing less of a headache?”
Luckily – yes, there is. It’s called “the Public Cloud”.
So what exactly is this Public Cloud?
The Public Cloud is a service provided (usually) by technology giants such as Amazon (AWS), Microsoft (Azure), Google (GCP) and others. It enables access to computing, data storage, network infrastructure and other IT resources for a periodic (i.e. monthly) fee or in a pay-as-you-go model, where you are charged according to actual usage. These amounts are usually very small compared to the cost of creating all equivalent infrastructure, although they can stack up over time. The Cloud frees its users from the need to purchase and manage physical infrastructure. All the fuss of provisioning servers, computing power, storage, network management, etc. can be vendor’s responsibility. Interestingly, although we are talking about “public” Cloud, such solutions are usually not only the best guarantee of reliability, stability and scalability, but also security.
…-as-a-Service: Cloud Flavors
The Cloud comes in many different “flavors” and each is best suited to address a give type of need. With big businesses’ love of acronyms, you can expect these to have sexy “nicknames” too and indeed, that is the case:
- First, there’s SaaS (Software-as-a-Service), where the vendor takes care of everything and you only get credentials to access the application via, for example, an Internet browser. This seems very convenient at first, but you may find that SaaS PLM solutions will force you to abandon your proven working methods in favor of pre-defined processes.
- Second is PaaS (Platform-as-a-Service), which serves ideally for the purpose of building a new application, for instance. In this case you’d only be responsible for the application itself and data you produce/store. However, you may fall into issues regarding vendor lock-in and potentially compliance.
- IaaS (Infrastructure-as-a-Service) rounds up the list of “traditional flavors.” It’s closest to virtualization, as it is up to you to manage nearly every aspect. You only don’t have to worry about stuff like operating systems, servers and storage. This requires quite a lot of hands-on attention, so it’s not for everyone, but done properly can be an effective way of hosting a PLM system.
- FaaS (Function-as-a-Service) is a method of executing code without any management of the infrastructure whatsoever. It’s perfect for atomic operations (and for building microservices), but not for complex things. And PLM is, indeed, complex.
Luckily, you’re not bound to one specific “flavor” and you can “mix-and-match” certain models for particular areas of application within your PLM environment. If done properly, this could bring you the peace of mind of SaaS, while maintaining IaaS-like control.
Cloud & PLM
If done properly, the combination of PLM and Cloud seems to be the perfect solution to ensure that the system works faster, more efficiently, scales, and is easier to manage and maintain.
Traditional PLM systems moved to the Cloud do not differ much from their traditional counterparts. They allow BOM management, workflows, change management, CAD data management and so on. Moreover, it’s likely that Cloud versions will even have the same user interface, so for many end-users, the difference might not even be noticeable. There is a significant difference in the underlying infrastructure, which also offers new strategies and models for system management.
There are also systems like Oracle’s PLM Cloud or Upchain Cloud PLM, which were designed “from ground up” to only work in the Cloud and may differ from your traditional view of a PLM. Others, like PTC’s OnShape, are mainly focused on 3D data and have limited PLM functionality.
Cloud-based vs. on-site solutions
Implementing PLM in an organization is a complex process which requires effort and knowledge to ensure efficient operation of this tool at a company’s premises. Looking from a traditional perspective, an IT department might have to do the following to launch and run a brand-new PLM system for an organization:
- prepare the server room, with specific air conditioning, fire safety, power backup and other requirements
- place physical servers and other equipment (mass memory, network, etc.) in the server room, connect them to the network, and then install and configure all required software
- configure the PLM system, including installing a database and required third-party software (e.g. Java runtime environment)
- set up PLM backup strategy, run backups and ensure that these copies actually appear
- manage and maintain the database
- manage and maintain the PLM system (patches, critical updates, version updates, upgrading the system to newer versions, development and maintenance of any configuration or customization specific to the needs of the organization, and so on).
If this seems like quite a lot of work to you, then you are correct. It is. Probably even more so than you imagine.
Is it worth to move your PLM to the Cloud?
The answer isn’t as simple as “yes” or “no.” Multiple factors must be considered. In general, however, companies which did commit to the Cloud report numbers like the following:
Please note that those numbers aren’t at all focused on savings on infrastructure, IT and so on. That’s a completely different category and I omitted it here on purpose, as it can vary greatly from one company to another.
If these seem too detached from reality (or too hard to translate into real dollars), please feel free to reach out to me or the Cloud Team at TTPSC for a free assessment of your environment and a personalized estimation of costs and savings.
Get More Business Value with PLM Cloud
Cloud PLM Automation
The phenomenon of the Cloud is that it was essentially designed with automation in mind. This means that everything you manage in PaaS (like application and data), or in IaaS (applications, data, runtime, middleware) can be automated and behave like in a SaaS model. Thanks to the Cloud the need for mundane manual tasks can be eliminated (or at least severely limited), as most things can happen without human interference while providing you full control over all environments whenever required.
Automation may require a lot of initial effort. However, it brings tangible benefits: once you do it, it takes 15-30 minutes to launch your new Windchill instance with full configuration and have the latest software build deployed. Compare this to a traditional (manual) process that can take days (or weeks at worst) and the benefit is clear.
Safety and Security
It is important to be prepared for a crisis. Every system fails, and the Cloud prepares you for that eventuality. The Cloud offers tools for automating system health checks and initiating action in response to issues, such as restoring backups or spinning up new instances, so disruption to your business is mitigated to even negligible levels.
Another convenience is the ability to create a centralized log that automatically notes all things done with/to your systems, so you can monitor essentially everything around the infrastructure. It’s not only useful from a maintenance point of view, but even more so for security purposes.
Myths about PLM Cloud
The internet is full of “wisdom” about every aspect of our lives, and it’s no different for PLM, the Cloud, or a combination of both. Often, would-be experts speak out, but their intention is simply to sell you a product, a solution, etc. For this reason, there are many beliefs and opinions which have nothing to do with reality. Obviously, these should not be followed when making decisions. While we won’t be able to address all of them here, here are a few examples:
It’s a common misconception that SaaS is all pleasure and joy. Yes, this model does not require us to configure a backup strategy, apply patches, etc., but simply to pay and use a system. An issue will arise, however, when you want to make a change to the system which is out scope of what the vendor offers as “out of the box” – which is usually quite limited. It is then that you may start wishing you had full access to servers to take care of this trifle yourself.
I’ve also encountered opinions that On-Premises PLMs are usually so over-customized that they are nearly impossible to upgrade. These voices suggest choosing SaaS as the only remedy. The truth is that such problems only happen when poor working methods are applied to developing your tailored PLM solution.
To sum up
The Public Cloud is a collection of future-proof technologies and services. No surprise, then, that we’ve seen a rapid increase in the demand for these solutions. Cloud has seen significant adoption, especially in recent months, when work-from-home strains companies’ networks. Many have already decided to migrate to various Cloud-based solutions and they are reaping benefits, regardless of their – or their products’ – level of maturity. You may not even be aware of it, but most of various business-to-consumer services you are using today are heavily Cloud-based. It seems, then that transitioning a PLM system to the Cloud, is the only reasonable decision, although sometimes it may be a challenge. Taking advantage of the new technologies and innovations brings noticeable benefits that translate into measurable value for most organizations. Having your PLM in the Cloud is a guarantee of data protection and security, scalability, as well as almost unlimited storage and computing power. However, before coming to a decision and starting any such transition, I suggest to carefully evaluate your existing infrastructure and all the variable values assigned to it: timelines, cost, resources and expertise. For once, you will certainly discover many ways to improve system performance, stability and cost-effectiveness in any approach you deem your choice. A trustworthy partner who will advise you, help you build a strategy and execute on it may be your most valuable asset in that process. After all, PLM never was and never will be a “one-size-fits-all.”