Fusion Product Roadmap Update

[prabakarm]

Just wanted to let you guys know that we have a dedicated subpage in our blog for roadmap.  We updated the Feb date and April content yesterday.  We will refine April content more and more in the next couple of weeks.  Our intent is to keep it live and up to date.  In addition to it we plan to blog about details of the problems we are working on and have a discussion with you guys.  We are really looking forward to it.

http://forums.autodesk.com/t5/fusion-360-product-roadmap/bg-p/79

BTW, check out Carl's blog on why we built Fusion if you have not already.

http://forums.autodesk.com/t5/design-differently/how-and-why-we-re-building-fusion/ba-p/5496355

Prabakar.

[cekuhnen]

I am confused now - where the t-splines to brep a duck tape shortcut and not a workflow?

this is number one reason why I settled with fusion to replace solidworks and teach it in my design class.

[Anonymous]

breaking definition of intended use is a way of life, and who cares what you call it as long as it is repeatable and gives expected results?

seriously, without the goal directed advice some may be expressing, which is not relevant to me,

I am also curious is there any update or expectation on when the mesh workspace may become available? 

I would love to use it to do some subtractive manufacturing of materials using complex shapes originating in .stl files...

I know @TrippyLighting has a lot of good things to say about the current capability,

and that is all good,

but the hints of what could be coming are pretty enticing,

and I'd rather we talk hopefully here in a roadmap thread about what will be,

than defensively of what we already have and whether it is good enough...

[colin.smith]

Hi @Anonymous

We are planning to release a preview of the new mesh workspace in July. 

Here is a link to a tech preview video that was done last year.  The UI is obviously out of date but the functionality will similar.

https://youtu.be/nQr20fTQS5M

Colin

[Anonymous]

WOO HOO!

[cekuhnen]

@colin.smith Hey Colin, what is the intendet usage for this mesh tool? Just basic clean up like in the video?

Or is it planed to also use then the imported mesh to map a new NURBS surface over like in Alias?

The TS worklfow (snapping to the surface) seems to be to be too complicated and labor intensive.

Just curious.

Claas

[colin.smith]

Hey @cekuhnen,

The short term goal is better overall handling of mesh bodies.  Clean up and heal mesh bodies as well as the ability to convert to quads for conversion to Tspline and Brep.

Longer term it will be tools for reverse engineering - the ability to guide the system for creating surface bodies (planes, cylinders, etc). 

Until we have full NURBS surface support the TSplines will cover the more organic shapes. 

Colin

[cekuhnen]

@colin.smith Thanks for the quick reply. That sounds like a reasonable approach.

The software "instant-mesh" is incredible at remeshing a body with quad faces.

Not sure if you guys looked into it.

Best,

Claas

[colin.smith]

We have some new mesh quad converting technology that hasn't been released yet.  Currently Memento uses quad mesh converter that was taken from Mudbox.

The new tech is more robust and should give better results.

[TrippyLighting]

That'd be very interesting to see!

[PhilProcarioJr]

@colin.smith

I really hope your team puts as much effort into developing the new mesh tools to handle object creation. Currently it looks as if the focus is on fixing existing mesh models and converting them. Although this is great, its usefulness will be very limited for broad projects. Right now it seems like an advanced version of meshmixer, but if you can't control the flow of the mesh its usefulness will be very restricted. The biggest hit T-Splines is taking right now is how hard it is to create surface flow to achieve an accurate representation of your final surface. I would still like to see core modeling tools from apps like Maya built into fusions sculpt/model environment. This to me seems way more useful then just a mesh repair setup. Those tools would be a perfect fit for creation/conversion to T-Splines with a CAD model being the end goal. The tools would work fine because in order to be in a sculpt environment history is off. For instance using scan data, most affordable scanners are junk, the surface quality is terrible. So its more about how do I end up with an accurate CAD model from an imported mesh (that's scan data) because getting an accurate CAD model from a stl or obj that is an exported CAD model to begin with is easy. Another thing that comes to mind is Sub-D tools for working with mesh data. Maya can easily covert any Sub-D model to NURBS giving you a perfect surface or solid model. Obviously I don't know what the end goal is for the toolset I just hope the team is looking at the big picture for the end goal of these tools. The current videos are inspiring but I'm more interested in seeing Fusion's end goal for these tools. I understand that it will be quite a while before we see that end goal as it takes a lot of time and effort to develop tools like this.

Either way I can't wait to see what the team has done with the tools.

[colin.smith]

Hi @PhilProcarioJr

Thanks for the feedback - useful as always.  The goal of the mesh environment is much more than just mesh mixer type functionality. Longer term we want to provide a solution for working with meshes from various sources - scans, photogrametry (Memento), and generative (geometry optimization, lattice, lightweighting, dynamo, etc). As well as providing tools to move from mesh to TSplines and Breps.  That will include clean up and reverse engineering tools as well as mesh triangle to quad tools so that Mesh can be converted to T-Splines.  We have multiple technologies in house and we are currently working out a plan to bring them in to Fusion.  It is a long term project that will phase in to Fusion.

Your point on the Mesh toolbox is well taken.  If there are tools in Maya that you use regularly for mesh prep/repair I would like to know about them.

Thanks

Colin

[cekuhnen]

@colin.smith I am really curious how your teamm and devs wills olve the reverse engineering process based on the scan data.

Till today I find with the tools available the best practise is to combine scan data and hand measurements to rebuild a new quality

surface by hand.

To me this seems to be technologically a tricky thing regarding the precission of scan data and what BREP can offer.

[RandyKopf]

Hello @colin.smith

I think this is one of the better aspects of Fusion 360 and Autodesk. That you are willing to actually say what you are working on. Too many companies leave you in the dark and you get what ever is served and only when it is ready.

Thank you for being transparent!

[colin.smith]

@RandyKopf

Thanks for your kind words. Although plans and priorities can change and sometimes items we talk about don't happen when we original thought - we try to be as transparent as possible.

[Anonymous]

I'd like to point out that a number of meshies are hanging out at http://forums.autodesk.com/t5/design-validate-document/stl-file-import-for-sculpting-edit/m-p/626537... and Phil has been explaining in great detail his tools and workflow there to the great advanatage of all of us. 

SInce we're explaining which aspects of the mesh toolbox are of interest, I would like to contribute a set of use cases which are quite different from those of Phil. 

He is interested in measurement accurate translation of primarily medical or engineering imagery;

I am interested in found object workflows where I might be able to take scans - either my own using an occipital scanner or those found on other places like thingiverse - of objects of antiquity,

and replicate those in various materials using subtractive milling.

My current rendering platform is an inventables x-carve;

my hope is to render images using that into cast off items like left over hardwood flooring and blocks of firewood.

My current challenge then is that the 8k face limit in Memento and the 10K limit generally in Fusion for mesh import result in a general inability to bring in complex shapes;

the inability to easily translate the meshes into quads without a fragile multi-app workflow where I need to learn several complex user interfaces (think Blender...) is a real impediment.

as well, most of the meshes have 'issues' and therefore will not convert to solid easily if at all, if I can get them into fusion with any remaining semblance of their original geometry.

in addition it is not clear to me how to generate a workflow which will allow me to scan irregularly shaped objects for use as stock,

(think block of split firewood and the desire to carve into it) and be able to use that as a basis for a toolpath with any sense of target surface accuracy.

I know these are not usual use cases and I'm sure some others may see these as frivolous, 

but they represent challenges I am trying to understand how to solve, 

and I am hopeful your mesh workspace will have tools which help me to do so.

Thanks for the opportunity for this conversation.

when the release comes out, will we get a separate mesh forum space so we have a place to gather and discover how to do this sort of thing together?

[TrippyLighting]

Your use cases might be different from the topics discussed in the thread you are referenceing, but Phil and others have been explaining a bunch of other workflows in a bunch of other threads.

The use case you are desctribing not particularly so unique or and not that different from what other users have tried and succeeded at. Perhaps with the difference of defining cracked Firewood as stock in the CAM environment. I don'd use the CAM environment, so I cannot comment on it's feasibility for that purpose.

STL import and mesh import have been discused rather excessively here on the Forum and a range of workflows have been proposed and used. Someties that simply includes other applications such as Instant Meshes, Memento, Meshmixer, TopoGun, Blender etc. I personally use Blender for several thing that I cannot do (yet) in Fusion (Heck, I could esily make a feature length video explaining al the things in Blender that would enhance the current Sculpt work space). But Blenders UI is not for the faint hearted 😉

Sometimes there is simply no way around using a specialized application. That does not necessarily mean, however, that it creates a fragile workflow. It would certainly be nice to see some more mesh abilities in Fusion if only for the ability to seemlessly work with meshes with a higher poly count. But in that case I am talking about a factor of 100 rather than 10. The 10k limit is extremely low and hasbeen a very strong limitation for so many projects that work with scanned data.