Welcome to the Onshape forum! Ask questions and join in the discussions about everything Onshape.
First time visiting? Here are some places to start:- Looking for a certain topic? Check out the categories filter or use Search (upper right).
- Need support? Ask a question to our Community Support category.
- Please submit support tickets for bugs but you can request improvements in the Product Feedback category.
- Be respectful, on topic and if you see a problem, Flag it.
If you would like to contact our Community Manager personally, feel free to send a private message or an email.
Dream CAD
Now it is the best time for us, the users, to motivate the new CAD system to hold our best needs. OS should be more then just SolidXXX on the cloud.
I think it is an opportunity to throw our wildest ideas and let those brilliant people find a smart solutions.
I think it is an opportunity to throw our wildest ideas and let those brilliant people find a smart solutions.
Gal Raz
Tagged:
2
This discussion has been closed.
Comments
**Dream CAD wanna-have 20140918**
Direct editing on the model in the graphics area. Using a sophisticated all-in manipulation tool.
What about a Stormboard brainstorm session? Here's mine to join...
https://www.stormboard.com/invite/105755/patent26
Storm ID: 105755
Storm Key: patent26
Dries
Leverage CAD data in a product configurator that sales representatives or clients can use to visualize and configure (select, assemble...) products. Have an automated way to generated bespoke poduct information (drawings with basic dimension, standard views...). Also: have to ability to download a configuration in neutral formats (STEP, IGES, DXF...).
Ideally, the configurator should always reference the latest released versions of all products.
I get a lot of requests from our clients, from our sales guys and from service bureaus (rendering, BIM...) to generate STEP's, DXF's... of our products. If I had to pick ONE area in my workflow that is too disjointed, too error prone, too time-consuming, too hard to keep track of versions, it would be providing appropriate 3D data to external parties.
Dries
I've looked at GrabCAD Workbench and I see how it can be useful, but it indeed doesn't directly address the versioning issue. Whatever is inside Workbench is completely separate from your main CAD. Same thing for any other 3D catalog portal, I guess. I see a path where Onshape could do both: CAD creation & 3D publishing, all tightly linked.
I agree, let the OS subscriber decide what and how geometry is exported for different needs (workflows?). Let the client decide what he needs to download.
Dries
"One file, hosted platform, lots of exports and viewing options chosen by the end user." THAT'S ONSHAPE.
(with one exception, there are no files in Onshape).
Onshape turns the export/translation/send-to-vendor activity into a vending operation. You simply share an Onshape document containing the (3D) information with your vendor or partner and THEY control what they need; all governed by the credentials you provide them (Edit, View, (more coming)). Your vendor can work in Onshape creating their designs variations on branches and workspaces, and download in their own desired format (X_T, IGES,STL, ... whatever).
An example: you model a part to be molded and share it with a vendor. The mold shop (the supplier) signs into Onshape, quotes the job and you send them a PO. The supplier then creates a branch (from the versions graph in Onshape )of your design; adds draft and makes a few other tweaks related to the mold process. Then they download your model in the format they prefer. They make your part and send you the part & the invoice. You like it and send them a check. AND THEN, here is the big part, you unshare the document with them and they no longer have access to your original model. Try that with GrabCad Workbench.
All activity and changes are tracked in Onshape. Versions, part numbers, etc are maintained. Even the neutral files or tool-path files can be uploaded back in Onshape and assigned to a version.
Bottom line.... let your vendor (the consumer of the data) access your original model in Onshape then THEY select the type of format they need.
BTW, have any of you experimented with the version graph yet?
The time consuming factors for us are simply providing 3D and 2D data to non specialist users. For example, exporting 3D data to Revit, requires saving as ACIS, and exporting just the external surfaces as a lump. SolidWorks has means of defeaturing parts to simplify exports, and to be honest I would likely do the same in OS creating an exportable version.
The big holes in the Workbench experience at this time are Partner Spaces are not secure...it is an open link. What is needed is an open link (no log in) combined with a password WE issue to the supplier. This way the system knows who accessed it and when. The problem with GrabCAD is that it is in the interests of getting users to set up accounts. We are only interested in getting secure connections to our suppliers and customers. Partner Spaces are useful but not a long term solution for us. I doubt the situation will alter under Stratasys.
And yes, I am dabbling in Versioning when I have time to experiment! Role on the iPad version!
Thanks for the details, I am familiar with Workbench and the limitations but really appreciate your real-world experience. The open share is certainly an issue. Your suggestion of a supplier portal where you control the access (even the OS access) is an interesting one. Of course, with Workbench, the supplier does have your original data even after you delete the Partner Space. Not necessarily so, with OS.
When you say "The concept of supplying unfinished part data to a supplier is alien here....". Do you mean that you would always provide your supplier a completely machine ready model? With draft, or parting lines or bend radius or whatever their process requires? I assume then that you are not sending a CAD model but rather an IGES or other format.
Do your suppliers have CAD? If they had OS would that change your process significantly?
The supplier will always have data sufficient to make the tools or parts, regardless of platform. Contrary to what most CAD vendors believe or promote, the global world of the SME is based on distribution of STEP or IGES, and that means data sufficient to make parts. Cannot see how Onshape or any other system will alter this? Once the data is out there and into a CAM system or other CAD system that is it. It is out there.
WRT production data, yes. We supply ready to go data with draft, split lines, etc. so do most UK design companies and manufacturing companies. I've heard of this strange beast of designer who sends out unfinished data for others to finish off but not actually met one :-)
Finally, yes all our suppliers and customers have CAD. Even the smallest companies do these days. A straw poll shows CATIA, SolidWorks and Pro/E are the most prevelant. Some use NX, one or two SolidEdge. I don't see how if they had OS it would change our working relationship that much. The fact is once a CAD system is in place, few want to change. Even all the recent scare tactics from Siemens over the future of SolidWorks has not dented the numbers in the UK (from what I see). Sure a few switched, or rather, they bought licenses on the "pay subs only cross grade" deal, but most of them did so to buy in extra CAD seats!
The only way to get new CAD in place is to sell it at a price that cannot be ignored AND offer the same or better modelling and deliverables functionality as the incumbent system. By deliverables I mean outputs like drawings, 3D data, data controls etc.
If you cannot model it, you cannot do anything else downstream.
And like Kevin says, sometimes we need to deliver translations, visual product envelopes that are pretty far off from being CAD. I’m thinking about Revit (we have a HUGE demand for this in the area of BIM!), OBJ polygon exports (rendering, 3D in non-CAD envirmonment…) etc. So I’m wondering how Onshape will handle exports from native CAD to not just CAD-compatible formats (STEP, Parasolid, ACIS), but also to completely different 3D representations (ACIS with external faces etc.). And how we –as Onshape users- can have full control over exports. I’m warning you guys… I’ll kill for the iPad version! :-)
THAT will be the true game changer. I agree with most points.
Resistance to change in CAD infrastructures is a real (cultural) problem. The problem obviously grows the larger the design teams and the more intricate ERP/PLM/… dependencies get.
Yet still, companies like Daimler Mercedes undertake MASSIVE multi-year projects to switch from CATIA to NX.
If the alternative is really really good, people will always consider it. (Or at least they should…
I think for Onshape a few conditions will have to be met to make this happen:
- really solid CAD creation and management tools (at least on par with midrange solutions)
- extremely good user-experience, which should reflect in: easy learning, clear interfaces, efficient workflows, solid help resources
- ’pervasiveness’, as in: consistent experience on all platforms
- somehow attract users that would otherwise not do CAD. By doing CAD differently, making it more intuitive. Not just by doing CAD in the cloud.
- an attractive, flexible and CLEAR (!) pricing model
With Onshape being built from the ground up -cherry picking from the latest and greatest technologies in servers, databases, CAD...- there is a huge opportunity to do something remarkable in this (arguably) stagnant market that is CAD.And I'm not just saying that to earn badges!
But these are just my ideas...
What do you think?
Dries
Great stuff and I completely agree re: resistance to change is a force which can be hard to overcome. Please keep the ideas coming.
You mention how important it will be in the future to engage the non-traditional CAD and "not just CAD in the cloud"...... I am thinking about the company supply chain; In your experience do the SLM people generally produce and send the the required CAD data (IGES,STEP, etc) to the supplier or does the Engineer generally provide that data?
iPad.... are we doing an iPad app?
We see this as a service to our clients.
Therefore it's people with direct access to the CAD software (meaning: designers) who generate the exports. We don't have CAD assemblies with every possible product configuration and variation. So these assemblies are built on request.
The main tools used in our office: SolidWorks for anything CAD & STL, modo for OBJ, KeyShot for images and animations.
It would be great if sales or customer service could do these custom assemblies and exports themselves, using the base CAD that the designers provide.
Dries
Fully hybrid surface modeling.
- subdivision surfaces, which are: editable, directly interpreted as NURBS bodies
- Parametric freeform surface modeling
- Explicit surface modeling (CV editing, surface matching...)
I realize some competitors charge $$$ for this functionality, especially with regard to the last bullet...But it doesn't need to cost $$$. I use Rhino together with Autodesk Shape Modeling Plug-in for explicit surface modeling. Very powerful capabilities for a modest cost.
Dries
the change:
I prefer analytical faces over planes because they are easier to control. Wish I could make them transparent and they do clutter up the feature list. Hoping feature list gets cleaned up soon so it's stops being unmanageable.
2c...
As Kevin , we are a small company (often just me !) , we tend to only release fully finished dumb files (eg x_t etc. ), rarely all the parts or a full even dumb dataset. This for external people to make... Often in China, or at machine shops all of who have there own cad systems. Even the word 'share' gives me fear if related to a whole product dataset - in fast moving consumer products copying is rife ( I hate alibaba ... Where I see many factories making poor fakes of our products).
But I guess the biggest market for CAD vendors are big companies.. Multiple seats of cad and much more 'design by team' ? ... So I am guessing we small co's by comparison are of less interest ? .... Although I would suggest LOTS of small business' using CAD is a bigger market .. (But more diverse .. Like trying to hurd cats :-) )
2nd to awesome, easy, and powerful surfacing .. As @DriesV suggested above,
..... I loved using catia's 'Imagine & shape module' on a huge automotive client's cad monkeys rig. The user had no idea of the power ( or the cost !! ) of his catia seat .....
This seems to be a typical result of the cad industries focus on big multi seat companies.
But less interest in smaller cos where there is often more advanced use of the cad, but less money.
Thoughts ?
Mark
How about a physics engine option?
Gravity, surfaces, friction as a start. I suppose there would be a need for simplifications when dealing with close fits - maybe tied into the ISO standard fits system?
Perhaps this could also tie into inferred or automatic mate connectors? For example, a rotational mate would be implied when a shaft is inserted into a bore. A bolted connection (more than one bolt) would imply a fastened mate. Maybe this could be used as a temporary state to help apply the high-level mates, then once these mates are established the computational requirements would be accordingly reduced (once a mate was applied, the physics engine could ignore that connection and treat it as being resolved).
Significant compute resources would be required - but like rendering this might be addressed with the OS cloud architecture.
Companies like LAGOA are trying to do this, but I think their business model isn't sustainable.
I would rather pay for rendering hours as I need them. "Choose your rendering package: 1 hour @ €7, 5 hours @ €30, 10 hours @ €55..." Rendering time over? Buy a new package.
What do you think?
Regards,
Dries
-J
My dream cad would have all repeating things automated or semi-automated for creating/learning new slightly different automation for next similar task. Ending up with completely automated cad which can read your mind and models just appear on your account.
- be based on a concept where the CAD provider's needs continue to pull them in much the same direction as the user's needs (so far, a tick)
- never need upgrading (big tick)
- be simple to drive for simple jobs (semi-tick)
- be enjoyable to use (semi-tick)
- be forever striving to provide useful capability at the margins (to soon to judge, but all the right noises)
- pursue a coherent and durable vision to counterbalance the "race to the bottom" which a focus on ratings tends to foster over time (too soon to say - it may be that the unparalleled ease of delivering feedback means that much of it is insufficiently filtered at source)
- make it easier to work flexibly and collaboratively (BIG tick)
---
In due course (by public launch time): my dream CAD documentation would be brutally clear about current limitations (so that the user was not forced to continually experiment in an attempt to infer them). This was a VERY weak area for Solidworks for the first decade.
It's not happening consistently yet at OnS, (witness output of 2D profiles being limited to the most recent version of dxf, not mentioned in "Help") but that's understandable given the pace of improvement. Some of the necessary mechanisms -- feedback and responsiveness -- are firmly in place: the acquired knowledge only needs to be patched through seamlessly to the documentation team. (Limitations are generally more visible to users than developers)
My chief concern regarding the OnS model is this: Future proofing, and even availability, of live models is reliant on the continued presence and viability of Onshape online. I realise that most modern CAD providers, following Microsoft's lead, have found various ways to lash us to their apron strings. This leaves us almost as vulnerable as the Onshape business model does, and it's particularly an area of concern in the case of products with a very long time horizon.
Not every design becomes obsolete in months or years. Consider, for instance, the grand piano, the jackhammer, or the revolver.
Security is not just about protection against malicious third parties, it's about long term autonomy in respect of one's key business processes and assets.
Otherwise, I'm liking OS quite a bit at this point.