concurrent engineering workflows

[Frederick_Law]

https://forums.autodesk.com/t5/inventor ... 37#M834137

Very interesting

[mike miller]

"Collaboration" is one of those concepts that sounds great in theory but has a tendency to turn into "colloboration" in real life. I really have no interest, nor see any need for, a program that allows other users to edit the same model that I am working on. If they want to watch me or even, (God forbid) "help" me, there's always Zoom which has screen-sharing and multi-attendee control. Call me a hermit but I prefer not to work simultaneously on a model with other people, thank you very much.

I'd much rather see those R&D dollars go to other places.

[mattpeneguy]

IV-Concurrent-01.jpg
https://forums.autodesk.com/t5/inventor ... 37#M834137

Very interesting

Hey @Frederick_Law, why don't you make this suggestion over on the platform...it'll give those 500+ people something to do...they're obviously not fixing the platform itself...

[bnemec]

"Collaboration" is one of those concepts that sounds great in theory but has a tendency to turn into "colloboration" in real life. I really have no interest, nor see any need for, a program that allows other users to edit the same model that I am working on. If they want to watch me or even, (God forbid) "help" me, there's always Zoom which has screen-sharing and multi-attendee control. Call me a hermit but I prefer not to work simultaneously on a model with other people, thank you very much.

I'd much rather see those R&D dollars go to other places.

I don't know anything about it, but my imagination goes places like, what if there were no assemblies, no file references? So if the entire project is one model there must be a way for more than one user to work on that model unless you're in a one engineer job shop.
What if there were no CAD files even? What if the cad model being displayed was actually all stored in a database? All the features or the kernel instructions were stored in table(s)? Then would it be possible? Probably not with Parasolid as it's order dependent, what if there was a different kernel that isn't history dependent? Could the model still be deterministic? If so there could be multiple changes going on at once and rebuilding of the model would become asynchronous (multiple threads) something CAD users have wanted for years.

Like I said, I don't know anything and that could just be a bunch of hogwash I made up. To be honest I'm not interested in it either as it wouldn't work at all for how we use parts where I work. I too would much rather see our maintenance money spent putting finishing touches and polishing on the existing feature set. As Alin calls it the missing 5%.

[Frederick_Law]

He said Onshape already has it. Maybe worth a try and see what it does.

[bnemec]

He said Onshape already has it. Maybe worth a try and see what it does.

Are you talking to me? I'm not really in the mood for trying any new software right now. You guys go ahead and give it a go, I'd like to hear about it. I would consider using it for non-work stuff but my non work hours do not include anything CAD related.

[Frederick_Law]

Not targeting anyone specifically.
Actually with how things are going, I might get more "non-work" time to learn a few new things.

[MJuric]

"Collaboration" is one of those concepts that sounds great in theory but has a tendency to turn into "colloboration" in real life. I really have no interest, nor see any need for, a program that allows other users to edit the same model that I am working on.

I have a need for it quite regularly. When working on larger projects where there are multiple levels of assemblies at one point or another two or more people will be working on S/A that need to have something done with them in the next level up or multiple levels up. Problems as is are.

1) Only one person can have that assembly checked out from the PDM at the same time. If I need to add an assembly, change relationships in that assembly etc etc I have to wait until I can get access.

2) Same situation as above, but now someone has it checked out. I'm designing based on an older version. They check it back in, I look at it and the things I've been designing around are not different or what the last person added is in the way of what I'm doing.

3) Access for someone to see what you're doing, using etc or for you to see what someone else is doing. This is always a problem for continuity in larger projects. 10 people designing different aspects of the same project. Bob uses bearings from one company, Lisa from another, John from another and so on. If you are truly collaborative I can log on, pull up Bob's assembly and see what bearing, bearing family etc etc and use the same thing. I don't need to E-mail him, yell over the wall or anything else.

4) Project management - Sure you can have a Zoom meeting to go over stuff and "Share", but that's wasting time. If I'm a project manager and I can sit down, open up what you're working on and look at what you're doing and then make comments, leave examples drop in what I want or what I think is a better idea, there's no loss in time, no scheduling a meeting, nada. The other person never stops working or doesn't even have to be present, you don't have to set up a meeting and so on. You can do this 10 times a day if you want.

If they want to watch me or even, (God forbid) "help" me, there's always Zoom which has screen-sharing and multi-attendee control. Call me a hermit but I prefer not to work simultaneously on a model with other people, thank you very much.

I'd much rather see those R&D dollars go to other places.

Collaboration is a huge time saver. I 100% think it's a great idea and well worth the R & D dollars....Doing it well is not something most software packages are doing.

Edit to add: 5) Can't tell you how many times over the years I've changed a part that interacts with another part that someone else is working on. You either have to send them an E-mail, model or something so they can change their part or wait to change it yourself. It's often the case that when you go to change whatever needs to be changed that you find out it causes an issue with that part. Then you have to go back and change the part you first changed and do the entire cycle all over again. A truly collaborative system would allow two people to work on the same part. I could go in to that part and make the design time at the same time I was working on my part. I could see the issues, problems and the part would change in real time so that the person working on it could also be aware of your changes. You from a couple back and forths to none, no waiting and worse yet a much less chance of the change on the other part not happening at all because it gets forgotten etc.

[MJuric]

He said Onshape already has it. Maybe worth a try and see what it does.

I've been using Onshape at home for a couple months now. Not really spent enough time with it to even consider myself "functionally literate" with it. I would like to see the "Collaboration" on it used though. I'd just like to see how it works.

[MJuric]

I don't know anything about it, but my imagination goes places like, what if there were no assemblies

I think this is the direction Onshape is going although admittedly I have no idea what I'm doing with that software. I do know that they lean far more toward "One file many parts". It seems alot to me like SW weldments for everything. Their "Extrude" has Add/Subtract and "New". If you extrude a sketch and select "New" it creates a new part.

I'm still struggling with what the difference is between a "Part" and an assembly in Onshape. They have differences but they seem very similar to me in actual function. Then again I have not done anything really "Assembly like", moving parts etc.

[jcapriotti]

Onshape has the ability for multiple people to change the same part/assembly as a branch. Then you can merge the branch later, although you may have to resolve conflicting changes.

So for example,
User 1 creates a branch and adds a hole.
User 2 creates a branch and modifies the size of a cutout.

Each branch gets merged back into the main workspace and updates the model at different times. As long as the hole added doesn't interfere with the cutout changes, there are no model issues, both set of changes show up in the main model version. It's a very cool feature I wished I've had in SolidWorks since the beginning.

[bnemec]

Onshape has the ability for multiple people to change the same part/assembly as a branch. Then you can merge the branch later, although you may have to resolve conflicting changes.

So for example,
User 1 creates a branch and adds a hole.
User 2 creates a branch and modifies the size of a cutout.

Each branch gets merged back into the main workspace and updates the model at different times. As long as the hole added doesn't interfere with the cutout changes, there are no model issues, both set of changes show up in the main model version. It's a very cool feature I wished I've had in SolidWorks since the beginning.

This happens to us too much. So maybe I'm more interested than I thought.

[MJuric]

This happens to us too much. So maybe I'm more interested than I thought.

Do you mean you have two people changing the part at the same time?

[MJuric]

Onshape has the ability for multiple people to change the same part/assembly as a branch. Then you can merge the branch later, although you may have to resolve conflicting changes.

So for example,
User 1 creates a branch and adds a hole.
User 2 creates a branch and modifies the size of a cutout.

Each branch gets merged back into the main workspace and updates the model at different times. As long as the hole added doesn't interfere with the cutout changes, there are no model issues, both set of changes show up in the main model version. It's a very cool feature I wished I've had in SolidWorks since the beginning.

I'll have to play with that. I've seen the branch thing but didn't realize that's how it works.

How does it deal with errors. Say for instance a person puts a hole in and then someone moved the surface so that the hole is now in space? Does it just not merge or does it merge and show the error? Does it tell you that the merge has not taken place because of an error?

[MJuric]

Each branch gets merged back into the main workspace and updates the model at different times. As long as the hole added doesn't interfere with the cutout changes, there are no model issues, both set of changes show up in the main model version.

I'll have to play with that. I've seen the branch thing but didn't realize that's how it works.

How does it deal with errors. Say for instance a person puts a hole in and then someone moved the surface so that the hole is now in space? Does it just not merge or does it merge and show the error? Does it tell you that the merge has not taken place because of an error?

When does the merge take place, IE can it take place while another person is using the model?

[bnemec]

Do you mean you have two people changing the part at the same time?

Pretty much, they may not both have the file open at the same moment in time but we have ECRs that overlap. I call it leap-frogging. One change (ECR A) starts but will be a long change to implement in the meanwhile another change (ECR B) comes through that changes the same file and must be implemented first but not implement the changes made in ECR A. So we would merge change B, release the part for production then merge change A later. Parent (where used) assemblies also get sucked into the mess.

[jcapriotti]

I'll have to play with that. I've seen the branch thing but didn't realize that's how it works.

How does it deal with errors. Say for instance a person puts a hole in and then someone moved the surface so that the hole is now in space? Does it just not merge or does it merge and show the error? Does it tell you that the merge has not taken place because of an error?

When does the merge take place, IE can it take place while another person is using the model?

Sometimes you get errors like lost sketch relations you can fix. In some cases the changes in one branch will overwrite based on last merged like if two branches change the same dimension.

[jcapriotti]

Pretty much, they may not both have the file open at the same moment in time but we have ECRs that overlap. I call it leap-frogging. One change (ECR A) starts but will be a long change to implement in the meanwhile another change (ECR B) comes through that changes the same file and must be implemented first but not implement the changes made in ECR A. So we would merge change B, release the part for production then merge change A later. Parent (where used) assemblies also get sucked into the mess.

We live the same life it seems.......a messy one at that.

[bnemec]

We live the same life it seems.......a messy one at that.

time yet?

[jcapriotti]

time yet?

Always time.........maybe at the next SolidWorks World....err...3dxWorld.....or Swamp World...whatever they call the next one.

[Frederick_Law]

We are doing it on paper. Multiple copies in the shop and office. Depends on which one the boss had when changed.