Time for PLM (Windchill?) 2

[geesamand]

It's been decided that our Engineering team will be implementing a PLM system in 2019. 10 users. Moderately sized assemblies and no concurrent engineering. Three sites globally, one in UK, two in US, for now.

My last interaction with PLM was over 15 years ago in the days of Intralink. The software was buggy, greatly slowed down basic workflows, and required a huge server with significant administration time. And of course, once you take the plunge there's no backing out of it.

My first questions:
Is Windchill the only game in town?
Any advice for keeping things reliable and performance snappy? We do a lot of small design revisions and transactions and so check-in and check-out times are important.
Are there viable cloud-based solutions?

Thanks,

David

 

[3DDave]

Yes, Windchill is basically the only game in town. And if you didn't like Intralink then you really won't like Windchill.

The talent and servers to keep Windchill snappy won't be available to a 10 user group. Especially when compared to Intralink which was a relatively low overhead operation.

It's probably worth it to find a PLM system you like and dump PTC entirely.

 

[looslib]

Windchill is the bets 'fit' with Creo and you have a choice of Intralink or PDMLink flavors.
Windchill by itself is nothing more than a very large and complex toolkit built on either an Oracle or SQLserver database.

Intralink will work fine if you only want to do CAD vaulting, no workflow, no part structuring, etc.
PDMLink will provide the full suite of tools to do workflow from problem reports to change notices, change tasks, CAD vaulting and finally releasing the finished product revision.

Performance of any Windchill based system depends on the hardware. I have 2 dual hex core machines (12 processors, hyperthreading disabled) with 36GB of memory in each box. Disks are 100GB for the OS, 200GB for the application drive and 400GB for the vault drive. We have over 1.2 million files in Windchill/PDMLink. My current vault size is 172+GB and an Oracle exp dump is 17+GB.

Windchill can be hosted on the cloud by utilizing AWS. Check with PTC if this can be implemented across multiple countries.

Alternatives are available. Teamcenter from Siemens can do PLM with Creo data. Drawback is upgrades need to be coordinated between PLM and CAD because it takes them time to make sure new releases are compatible. (More time than it takes PTC to be sure Windchill and Creo are.)
There are some other lower cost vaulting packages, but I am not sure the volume is worth the risk.

It does take a lot to setup Windchill for use. It is not for the easily frustrated to build a system from an OOTB install. Even a PDMLink OOTB install is an empty shell. No users, no roles, no vaults. Workflows and lifecycles, revisioning, etc, all have to be customized to your company rules and processes.

I have been administrating a Windchill environment for 12 years and doing CAD admin work for close to 30.


"Wildfires are dangerous, hard to control, and economically catastrophic."

Ben Loosli

 

[geesamand]

I got some clarifications:
1) We will perform preliminary study in 2019. The aim is to put appropriate resources into the 2020 capital budget to *begin* an actual PLM implementation. After a quick chat with our IT director, he thinking implementing in 2021 is aggressive, even for a small company.
2) The actual scope of the PLM implementation is yet to be determined. It will probably involve a PDM system, but that's not the only / main thrust
3) We currently have poor product documentation internally. The answers to questions about our products exist on the engineering drawings and are not distilled into detailed internal product specifications. In addition, the how/when/why/who behind the design or design change is only tribal knowledge and poorly documented. This aspect of PLM is a huge opportunity for us.
4) Our Engineering Change process was moved from a home-built app (and a good one at that) to our ERP software. It's not a good fit in the ERP system and very much an option to move it to a PLM platform.

This reduces my stress considerably - I'm happy that PDM is not necessarily being rammed down our throat in the name of "PLM".

 

[looslib]

With 3 sites, one has to be the 'master' and host the primary Windchill and DB servers. Your other sites become replica servers and the get a small footprint Windchill installed to host the data locally. The replica servers are configured to send any changed or added information to the master site at set intervals. When I had a replica server in Huntsville, Al and the main in Waco, TX, we had the replication set for 4 hours. Data hosted on the master would be transferred to the replica when someone at that site requested the data. The local server would be checked for the latest iteration before going to the master. You may need to be careful if your 2 replica sites share data, as you could get a request between replication of something saved at site B needed at site C before it has been put on the master at site A. Your replication rules will need to account for this scenario.
Another to check is the robustness of your data communications link between sites. Windchill does not like latency!


"Wildfires are dangerous, hard to control, and economically catastrophic."

Ben Loosli

 

[3DDave]

PDM/PLM won't fix the "Why" or "How" questions. PDM/PLM mostly allows release control and traceability and file dependency management.

I found that a Wiki is the better answer for managing information for "How" and "Why" about product development. However it's not directly linked to documentation release. It does include "Who" automatically as it records who made edits and "When" as it records when the edits were made. A Wiki also allows determining what was changed as well, for those interested in looking. It may be that some will allow highlighting a passage and asking "When" that passage was originally included, but the one I have does not.

What I like about mine is that I can attach files to the pages, so that a given page can carry all revisions of a document.

Unlike most information management systems one thing that Wikis allow is links to none-existent pages. These links are identified and unmistakable. The useful thing about this is that document structure can be created from the top-down. When those pages are created, the previous links become active without having to go back to all the places they were created.

Some of this are handled by features in PLM/PDM systems, but only people with licenses an a willingness at any moment can enter that data; if it is necessary to make a correction to this alternate data after the item is "released" then it's a painful process to correct it and has no better traceability than the Wiki has. Painful = no one fixes it and soon no one puts the information in at all. It works as long as people want to work together and nothing works if they don't.

 

[Hallec]

There ARE other options, I know as I've set them up in the past. One of my favorites is Agile PLM (

https://reviews.financesonline.com/p/oracle-agile-plm/).

My least favorite is probably Windchill. With a small group of 10 or less, I might prefer to have IT set up some basic permissions, a cloud-based CAD directory structure, and mirror the drives off-site using Windows over Windchill.

I'm not a vegetarian because I dislike meat... I'm a vegetarian because I HATE PLANTS!!

 

[CWB1]

I've used Windchill but still prefer Teamcenter for ease of use and implementation, its probably the most commonly used today. My employers have used it for everything and we've had ready access to store everything from design review presentations to standards in it.