PLM Is Best Hosted, SaaS, or On-Premise . . . Let’s Sort It Out
Oleg Shilovitsky posted on September 25, 2015 |
Every PLM vendor claims to support the cloud, but exactly how can be a bit of a mystery.
Here are a few questions generating some debate among both vendors and customers:
  • How will manufacturing companies be able to adopt cloud solutions?
  • What are the right products and technology for PLM cloud?

Earlier this year, I published a PLM vendors cloud service comparison article. I saw it as a way to differentiate offerings and provide some guidance for PLM professionals and customers navigating the sea of cloud buzzwords such as IaaS, PaaS, SaaS, public vs. private cloud and even things like virtual private cloud.

I had the chance to attend a few PLM conferences over the last couple months. After watching many presentations and talking to attendees, I found that the discussion about cloud PLM differentiation today can be narrowed down to a comparison between PLM hosted solutions and PLM SaaS services. It’s not easy differentiating between these two types of solutions, and PLM cloud marketing can often confuse customers.

In this article I will provide a short description of both and offer some simple guidance about the advantages and disadvantages of these approaches. My goal is to help you make an informed decision regarding future PLM cloud strategies in your organization.

Everything Is Hosted

This statement is actually true. From a certain standpoint, if you think about software running outside of your company data center, it is hosted somewhere.

The first time hosted applications were introduced to businesses we called them application service providers (ASP). For many people, it is hard to differentiate between ASP-hosted solutions and SaaS applications.

On the surface, the difference between the two solutions (hosted and SaaS) might seem insignificant. Both will release a company from the need to operate hardware and will shift responsibility to the hosting providers. But that’s where the similarities end and the differences begin.

Hosted PLM Solutions

Hosted solutions use PLM software that is capable of running on-premise and make it available to customers from an isolated hosting environment. In other words, hosted architecture takes your PLM servers and moves them outside your company data center. 

Hosted solutions can use the services of any hosted provider with computing infrastructure and, in some situations, customers may own their hardware. A hosted solution can be also used with infrastructure as a service (IaaS) providers such as Amazon AWS or Microsoft Azure, using a variety of computing infrastructure.

From an infrastructure and product architecture standpoint, these solutions will be almost identical to on-premise solutions. The application and data resides in an isolated environment.

Hosted solutions ensure the highest level of security, a clear advantage of them that is not very much different from on-premise installations. They can use multisite redundancy, and they can ensure the availability of applications in case of communication problems across a global deployment.

For the last decade, PLM vendors have gathered enough experience to run existing PLM products via hosted computing infrastructure. The disadvantages of hosted PLM architectures are the same as the advantages. They require isolated infrastructure, they don’t scale elastically (unlike modern cloud architectures) and they can limit some of the customization and administrative capabilities of PLM software.

Cloud SaaS Applications

SaaS solutions are usually represented by an application, which is running on top of multiple instances of virtual computers and using shared services. The architecture of SaaS applications is often different from applications that can be used in on-premise and hosted cloud installations.

SaaS applications take full advantage of the cloud infrastructure to run across multiple servers. On-premise products cannot and require some changes to the architecture in order to do so. SaaS cloud solutions also use computing and data services elastically and have the ability to scale according to demand.

The main advantages of cloud applications are their ability to scale elastically and their optimal use of cloud infrastructure. From the standpoint of scale, SaaS applications have almost unlimited potential and can support better cross-application communication. Additionally, the integration among cloud applications with SaaS is much easier and the data exchange is simpler.

The disadvantage of SaaS cloud applications is the fact that you are totally controlled by vendors and application providers. Such solutions cannot be taken down and installed on-premise, regardless of the cost. 

The integration of cloud SaaS solutions can be another challenge because it might lack some of the data hooks and services available in hosted products. While the majority of enterprise PLM software is integrated using direct access to SQL database, SaaS cloud integration should take a different approach and use Web API for integration.

The Modern Cloud PLM Solution Landscape

On one end, established PLM vendors have decided to make their existing PLM platforms cloud-enabled. The next logical step is for them to host PLM solutions using hosting providers and IaaS infrastructure. 

One the other end, providers of new PLM solutions want to leverage cloud infrastructure in full to run SaaS PLM apps, leveraging the elasticity of the cloud and its economy of scale.

Right now, the only two cloud SaaS PLM vendors are Autodesk PLM 360 and Arena Solutions. These two vendors are providing software only available as cloud SaaS services. The applications work through a browser, including basic administration, configuration and integration. You cannot run these products from a hosted data center that is on-premise.

For the last three years, major PLM vendors have hosted their PLM solutions through a variety of methods. Dassault Systèmes, PTC and Siemens PLM make their PLM products available via the hosted infrastructure of IaaS providers and partners. Aras Corp, for instance, supports multiple options for hosting Aras Innovator from Microsoft Azure cloud and via hosting providers (such as TSystem in Europe). Oracle Agile PLM is also available as an Oracle Supply Value Chain solution, which relies on Oracle Cloud infrastructure.

So, can you get the best of both worlds? It is not clear if PLM vendors will be able to support cloud SaaS and hosted solutions as a single technological and product platform. Some of the products (i.e. PTC Windchill) are still available via public cloud as a shared service, but they have limited configuration capabilities. I haven’t seen similar solutions available from Dassault Systèmes and Siemens PLM, but they may be available in the near future.

So, what’s my conclusion?

Cloud PLM is still a complex world, with plenty of questions for both vendors and customers alike. With traditionally long lifecycles and increased competition, vendors will need to adjust their approaches if they want to claim their share of the market moving forward. 

Meanwhile, traditional PLM platforms will likely face challenges of scale and need to leverage the cloud economy in order to provide effective PLM solutions at a low cost compared to modern SaaS PLM cloud services.

At least, those are my thoughts on the matter.

About the Author

Oleg Shilovitsky is an entrepreneur, consultant and blogger at Beyond PLM, a leading source for information and commentary on engineering and manufacturing software. Shilovitsky has been building software products for data management, engineering and manufacturing for the last 20 years. In the past he was developing PDM / PLM software and worked for companies such as SmarTeam, Dassault Systèmes, Inforbix (acquired by Autodesk) and Autodesk. 

Recommended For You