Will “cloud” change the way we integrate PLM and ERP?

Integration has a very important role in PLM implementations. PLM is intended to manage design and engineering aspects of product development. ERP is intended to manage manufacturing resources, process and inventory. Therefore PLM and ERP are complementary. Most of PLM implementations I’ve seen in my life, required PLM-ERP integration.

At the same time, PLM-ERP integration is often the most complicated part. There is high diversity in the ways manufacturing companies manage data about product, bill of materials, parts, inventory and manufacturing processes. Outcome is multiple BoMs, product and item records and the need to synchronize information.

Traditional PLM-ERP integration is complex and never done out-of-the-box. It requires detailed definitions, data mapping and variety of data synchronization techniques. The last one is usually code effort that done by service provider or IT programmers. In some situations PLM and ERP vendors are offering integration tools, but because of different reasons such as cost and complexity of these tools, integration often end up as SQL hacking into two databases of PLM and ERP. Software vendors are not appreciating this approach, but usually face the reality of large implementation complexity and just live with it. In most of these situations, vendors would not jeopardize PLM deal by preventing customers to access databases directly. The result is high cost of maintenance and problem during upgrades.

Cloud technologies are simplifying IT and deployment. But, at the same time, cloud can create an additional integration complexity. Traditional integration code, including SQL often not applicable without direct access to databases in web environment. But cloud environment is still very complex. It contains PLM, ERP and many other systems and services companies are using. Few months ago, I shared my thoughts about how to avoid cloud integration spaghetti. One of the biggest dangers is that closed data paradigms and data duplication between silos can bring well-know data spaghetti from on-premise applications to the cloud.

For the last few months, I’ve been learning about what cloud PLM companies are doing to simplify cloud PLM-ERP integration. I wanted to share some of my thoughts about it

Autodesk PLM360

PLM Connect is a complete solution portfolio provided by PLM360 integrate business systems. First of all, it applies to PLM-ERP integration, but not only. Earlier this year, at  PLM Accelerate 2015 conference in Boston, Autodesk promised to integrate PLM360 with everything. PLM360 is using Jitterbit middleware for integration.

In addition to that Autodesk seems to be inspired by IFTTT -like tools announced “evented web” integration for PLM360. Read more here.

From the side of Jitterbit, it looks similar to traditional middleware. The fact it runs on the cloud doesn’t change much. But it has nice UI for integration and mapping. Also, granularity of REST API and ease of code can potentially make PLM360 / Jitterbit environment more efficient. Evented-web integration style has advantages, but it is not clear to me how effectively it can be used to synchronize data between PLM and ERP environments.

Arena – Kenandy integration

I’ve been learning about Arena PLM integration with Kenandy ERP. My attention was caught by the following article and Arena-Kenandy partnership press release. You can get some details about the integration by navigating to the following data sheet.

I spent some time looking into specific ways integration is done. Arena and Kenandy is not using middleware style integration. At the same time, both are supporting modern web based APIs to code integration behavior. Which allows the both solution to leverage service APIs on both sides for efficient and granular data integration. Arena and Kenandy is synchornizing data by transferring XML documents.

Administration console can show you status of data synchronization.

In my view, Arena-Kenandy is a modern variant of point-to-point integration with realization using Web services API. It makes code easier, but still requires implementation of synchronization logic between systems.

Razorleaf – Clover Open Integration Platform

Companies doing implementation services for PLM usually have high sense of urgency to work on PLM-ERP integrations. It is part of their implementation schedules. My attention was caught by Razorleaf announcement yesterday about Clover Open Integration platform. Read more here – Razorleaf Introduces Clover™, a New Open Integration Platform that Supports Any-to-Any Endpoint Integration for PLM Applications. The following passage provides some high level explanation about what Clover does.

“The Clover platform is a result of our long-standing experience in creating CAD/CAM/PLM integration endpoints,” stated Eric Doubell, CEO of Razorleaf Corporation.  “We now have created an industry standard application integration platform that has a flexible architecture and can scale easily based on its endpoint applications.  This platform helps our customers retain the feature sets they have come to rely upon in their application investments and allow for a more controlled migration path forward when upgrading is a requirement.  Making up-to-date data available across applications accelerates decision-making and process efficiency across the organization.”

Razorleaf is providing services for different cloud and on premise PLM environments. Learn more here. You can see on-premise and cloud systems including Autodesk PLM360 and Jitterbit. I’m still learning about Clover technology and platform. So, stay tuned for updates.

What is my conclusion? Cloud brings some limitations to integration techniques. Very often integration was done using direct SQL-code injections and batch processing. You cannot do it anymore in cloud-based / web environment. Web based APIs can compensate it, but it requires products to support granular REST APIs for specific operations. This is something you want to be sensitive to when choosing cloud PLM vendor. Web API can make cloud-based integration easy to code and implement. However, cloud integration patterns are still the same – middleware or point-to-point integration. Cloud didn’t bring anything new here. At least from the standpoint of systems I learned. Integration remains complex and requires planning and resources during PLM implementations. A note or PLM architects and strategists. Just my thoughts…

Best, Oleg