90% of CAE users have no strategy for data/process management?
Check out the results from a poll on the New Trends in CAE Simulation linkedin group:
I'll be the 1st to say this is not a statistically valid poll. It's a small sample size of 126. Also, the people responding all belonged to a Linkedin group called "New Trends in CAE Simulation." That implies a likely bias towards innovative ways of working that might not reflect the general population.
That aside, the results are interesting. As I'd guessed, "File/folder naming conventions" came out on top. I did not expect "SPDM/PLM" to come in with such a healthy second place showing... again, likely group mentality bias. Kudos to those of you admitting to "No Strategy." And, a special hat-tip to those of you fessing up to creating a bit of job security in the "External HD hidden in the desk" camp!
In my opinion, File/folder naming conventions are only a valid strategy for simple products and tiny teams. It may have been working for you ok up to now, but pretty much all businesses are seeing an explosion of product, team, and customer complexity. Combine that with workforce aging/globalization/nomadism/reduction pressures, and you have just a few of many compelling reasons to come up with a proper CAE process and data management strategy.
So, I'd say these results boil down to 29% SPDM... and 71% little-to-no-strategy. In a broader, statistically representative poll, I suspect we'd see more like 10% and 90%.
How did we get here?
Partially this comes down to priority and digital maturity. PDM/PLM penetration is much deeper for things like 3D CAD and xBOM management. That didn't happen overnight. Some of you have the scars. Most folks aren't going to consider CAE until after those areas are under control. Also, in the early decades of PDM/PLM maturity, CAE was far less integral to designing products. CAE today has moved well beyond nice-to-have if you want to be competitive.
How do we improve?
The understandable knee-jerk reaction is, "Let's just manage this with the PLM tool our CAD team is using." Some of you who've tried this also have the scars. It doesn't work. CAE data is too complex, deep, large, and voluminous to work well in vanilla PLM. That's why you need to look at a proper SPDM tool like Teamcenter Simulation, MSC SimManager, or ANSYS/Aras Minerva.
Just remember that getting control of your CAE process and data shouldn't happen on an island. You need a strategy for connecting this part of the digital thread to whatever PLM tool(s) are already in place for other parts of the organization. In the case of Teamcenter Simulation, that happens by definition if you are Teamcenter customer. But, even if you use another PLM tool, Teamcenter Simulation is easily integrated with most PLM solutions via data hub connection technology.
Regardless of SPDM vendors on your shortlist for evaluation, start to inventory these 3 areas:
- What CAE tools do we use (both homegrown and COTS)?
You'll want to ask the vendors if they have existing data models for these tools, and what it takes to create a custom connection if one doesn't already exist.
- What processes or standard work do we really use?
It's not just data management. It's process management, too. Take the time to document the repeatable steps and people typically involved in a new project. Part of SPDM is bringing this "Standard Work" to life in a visual, trackable, directed workflow. It's for sure important in large, complicated teams with multiple CAE silos. But, I'd say it's just as important for smaller teams with simpler products to capture institutional knowledge. Think about how transient employees are today. It's a huge risk.
- How much have big failures or rework cost us in the last decade?
How often do you re-run the same kind of CAE process with manual setup? Put some numbers and time estimates on automating that. Multiply by the hourly burn rate of your engineering staff.
What big field or test failures can you remember where the original CAE models couldn't be found... so, you just re-ran the whole thing again from scratch? What kind of revenue loss, penalties, or customer relationship damage did you incur due to slow response/resolution?
What do you think?
Share your thoughts and examples in the comments below, and/or join us over at the New Trends in CAE Simulation Linkedin group. BTW, I'm aware of the oddly redundant name :)
A Digital Native for the Manufacturing Industry
3 年Thanks for sharing the survey and your article! It is almost frightening to see how old-fashioned working with simulation results is. Even the innovative and agile thinking experts work with outdated methods. This shows the big advantage and the great potential for SPDM & PLM for simulation world.
Studio Engineer | Mechanical Design, Analysis and Optimization | Stellantis | NIT Trichy
3 年My opinion. Process management could help in routine analysis but May not help In development. FEA engineers now a days don’t only give confidence to designers but also develops their own concepts. Users may face challenges such as - CAD is standard for all functional science but FEA differs . For same CAD there will be different meshes I.e. mesh model for durability , safety and NVH in case of automotives. Even the mesh density changes from system to full assembly for same component. Creating separate process management for different functional science may not sound beneficial. - Important features can be captured from CAD to FEA, but FEA to CAD requires expertise and time.
Engineering Team Lead at Bradken, Lecturer - Karaganda State Technical University
3 年I think there's a space for a new poll corresponding to CAE users' companies budget. It's much easier to implenet SPDM/PLM system according to existing business process at middle or large companies with small changes of those processes instead of developing or redevelopng of business processes in a small companies which are have no any strategy working with CAD/CAE. That is too expensive solution to store your data into PDM or having a full PLM solutions for them.