Application to Epic Games Megagrants for the BlenderBIM Add-on

M by Moult on 14 Aug 2020, edited 27 May 2021 # +4 votes

I have recently been contacted by someone from Epic Games, related to Unreal Engine. One of the things we discussed was the Epic Games Megagrant and whether the BlenderBIM add-on should apply. I have drafted up an application below. I plan to submit early next week.

Prior to reading, please read the BlenderBIM Add-on Roadmap, which talks about where the project plans to grow, and the likely necessity of funding required to allow the project to scale and provide the support that commercial companies require. Pinging @aothms - who without his work none of this would be possible, pinging @yorik - who has been pushing BIM in FreeCAD long before any of this started, and pinging @theoryshaw - who has been designing buildings in a transparent manner in a commercial setting in a unique practice unlike the rest of the industry, and pinging prominent contributors @Jesusbill @duncan @bernd @stephen_l @carlopav @htlcnn in your respective fields.

Four things I take a hard stance on:

I want the BlenderBIM Add-on to be a serious tool used by the industry. I will do the work it takes to get it there.
Money is useful. All of us are spread a little thin right now.
If possible, development is done with real commercial output, with trained AEC professionals working alongside coders. I currently use the BlenderBIM Add-on in a commercial setting in my day-job, and therefore that's what (mostly) guides development. I do not want to invest developer time isolated from commercial output.
Everything must be done transparently.

Draft in next post.

M by Moult on 14 Aug 2020 #

The application is in two parts. First part:

Describe your project

The objective of the BlenderBIM Add-on project is to provide a complete free software workflow in an OpenBIM production pipeline in the Architecture, Engineering, and Construction (AEC) industry. This is the ability to work from early feasibility and concept phase, project coordination, detailed design, fabrication & construction, facility management and building maintenance, post-occupancy evaluation, and tear-down. The AEC industry is currently proprietary dominated.

The main software package is an add-on for Blender. It provides import and export support for the IFC (Industry Foundation Classes) schema: the ISO standard of 2D/3D data + semantic metadata exchange in multiple formats (IFC-SPF, IFCXML, IFCJSON, IFCZIP). It also provides import and export support of BCF (BIM Collaboration Format), the standard for design coordination in AEC. Apart from file IO, it supports the relevant workflows of authoring and analysing building models, including semantic classifications and properties, classification system integration (e.g. Uniclass), clash / collision detection, data cleanup, CSV scheduling, quantity take-off calculation tools and costing, 2D drawing and construction documentation generation, diffing files, construction sequencing, automated quality audits, COBie facility management, lighting simulation, structural analysis, energy modeling, and parametric constraints. Much of its functionality is also additionally provided as standalone headless Unix-style utilities, which are shared with other free software projects like FreeCAD, and XeoKit. Similarly, the underlying library IfcOpenShell which provides schema parsing is shared between many free software projects, of which we contribute to.

I have successfully (but painfully) used it to deliver a small 250sqm highly bespoke building from concept design to fabrication and construction - including all drawings, construction sequencing, and scheduling. The add-on is currently being used in some aspects of large infrastructure projects in Sydney, including quality auditing, clash detection, and facility management data preparation. The project has seen publication in AECMag (alongside an open letter from UK architects to current proprietary monopoly Autodesk), BlenderNation, and has started a growing community called OSArch.org of AEC professionals in various disciplines, looking to use more free software in architecture, engineering and construction. There are now 400 forum members in half a year.

In addition to providing features equivalent to (and exceeding) existing proprietary software, the BlenderBIM Add-on project aims to be different in two ways. First, they will be highly modular, allowing users to mix and match to create their own pipelines without lock-in to a platform / GUI / vendor tailored to their needs, and easily modify or extend the pipeline if it does not suit their needs. Secondly, the tools will uphold the highest standards possible of open data standards, exchange, and interoperability conventions, allowing us to integrate data across disciplines to make wiser decisions in the design and management of the built environment. It is currently observed that the lack of robust integration between disciplines is a factor in poor design of the built environment.

M by Moult on 14 Aug 2020, edited 14 Aug 2020 # +5 votes

Second part:

Describe your next steps and plan to develop your project

New features and fixes are being built daily in my spare time. The pace of development is such that a new release is made every 2-3 weeks. Regardless of funding, I see no reason for this development pace to slow down given my current and near future financial and employment circumstances. Development is guided by commercial use-cases and challenges presented by my day-job, as well as bug reports and feature suggestions by power users and trained AEC professionals in the OSArch community. The commit log shows evidence of this, with a large variety of tasks being tackled across many disciplines. I suspect this approach of responding to community needs (with a priority on those that are using it on actual projects, not experimenting) will continue unchanged.

However, with additional funding that is large enough to hire dedicated developers, I will invest in two known shortcomings as a high priority:

New features to rapidly do architectural and MEP (mechanical, electrical, plumbing) 3D modeling with parametric shapes in a user friendly manner. Blender lacks many CAD transformation features and one-click parametric geometry that is needed to model buildings.

The ability to create 2D documentation (i.e. drawings & schedules) to the standard required by commercial jobs. Current tools are slow, clunky, and not feature complete. Without the ability to create construction documentation to a high standard, no free software will be considered seriously as an authoring platform.

There is no shortage of other things to work on (see commit log), but the two shortcomings identified above are likely to have the highest impact on the AEC industry.

Without enough funding to hire dedicated developers, I am exploring other options to use potential funding, including:

Partnering with non-for-profits to use free software for designing and constructing buildings for the benefit of society, with funding used to subsidise architect / engineer / building costs. This work will benefit the BlenderBIM Add-on, but will also benefit other used free software projects.

Partnering with academics, who are exploring using free software for research projects in the AEC industry. The underlying library used by the BlenderBIM Add-on and FreeCAD, called IfcOpenShell, was started in academia, and academics are usually involved in developing OpenBIM standards which are vital for industry interoperability.

Sponsoring membership to buildingSMART international, the non-for-profit organisation responsible for developing OpenBIM data schemas, including IFC, BCF, MVDs, bsDD, and the OpenCDE spec. This would help give a "voice" to the free and open source software community in developing open standards for the AEC industry. Membership is currently paywalled and given priority to large proprietary vendors.

Not using the funding for the BlenderBIM Add-on, but instead for other free software in the AEC space if we decide that their need is greater than ours. The AEC industry is huge, with much more than one tool involved. This decision can be made with the help of the OSArch community - which encompasses many disciplines and tools in AEC.

Leaving the funding untouched, but in a community fund that may grow with other contributors until we can do something productive with it, or be saved until a rainy day when devs need funding, or until a better opportunity presents itself when the project is more mature in about a year given the current rate of development (e.g. a project wanting to use free software comes along, or an enthusiastic developer appears, etc). I hope that as the project grows, I will, along with the rest of the OSArch community contributors, be equipped financially to ensure that AEC firms can be confident in the tools that we develop. Funding will help gain credibility among AEC firms who are used to proprietary tools and do not take free software options seriously.

No development / spending decisions will be made in isolation. Development is already done transparently based on the many suggestions of the OSArch community - they provide the insights, feature requests, and bug reports of many varied disciplines that need to work together to create the built environment. Any spending decisions will also be made with their consultation and support. This is a large, complex industry issue which is being tackled and needs the community support.

Cheers. Hope it resonates. And if it does, I hope it comes to fruition.

R by ReD_CoDE on 14 Aug 2020 #

I didn't read the whole, however, I don't know about IFC export, but Unreal natively support IFC import

Also, I don't see anything about the problem you try to solve? And the solution? And the final product?

R by ReD_CoDE on 14 Aug 2020 #

Have you seen Unreal projects in the AEC field?

As said before, you don't know anything about AI, automation and control, and XR and I hope you be able to find funds

E by edwinguerra on 14 Aug 2020 # +1 votes

Beyond this, now is a good time for all OSarch idealists here with means to do so to support financially this project. https://liberapay.com/dionmoult/donate I just did and hope that more will join!

M by Moult on 14 Aug 2020 # +4 votes

A point of clarification I should mention: the grant does not need to be related to Unreal Engine, Epic Games, or the game industry. If awarded a grant, I personally have no plans to invest it in integrating the BlenderBIM Add-on with gaming and Unreal Engine, as I personally do not see those integrations as currently high priority / high impact to the AEC industry. I think others would agree based on the feature requests seen in these forums. In my discussions with the representative from Epic Games, I have clarified that the BlenderBIM Add-on has little to no relationship to Epic Games directly (I was actually quite surprised to be approached, as I could not see the benefit to Epic Games). Both they and I are also aware that Unreal already supports IFC import. The impression I got from the conversation with them is that this seem to be a non-issue.

@ReD_CoDE the grant would be used "to provide a complete free software workflow in an OpenBIM production pipeline in the Architecture, Engineering, and Construction (AEC) industry". The problem is that currently there is no complete free software workflow. The solution is to continue building tools like the BlenderBIM Add-on. Tools like the BlenderBIM Add-on, along with its various utilities (IfcClash, IfcDiff, IfcCOBie, etc) is the product. Sorry if there was a misunderstanding - this has nothing to do with UE, AI, or XR.

@edwinguerra Thanks so much! I really appreciate it! For those reading, keep in mind there are also many other projects out there in need: @aothms is alone handling the core library for so many projects, and even projects like The GIMP only have about 6 main devs who do all the work (I heard in a GIMP talk somewhere). It's pretty wild.

Also to be transparent about my current non-labour costs:

40EUR a month for server to run OSArch forum / wiki, BlenderBIM.org website, build server
20AUD a year approximately for the BlenderBIM.org website domain

In short, a very, very small amount. The donations received from Liberapay actually already completely cover all of these costs.

D by duncan on 14 Aug 2020 # +2 votes

Very interesting. One things that interests me is "Partnering with non-for-profits to use free software for designing and constructing buildings for the benefit of society, with funding used to subsidise architect / engineer / building costs. This work will benefit the BlenderBIM Add-on, but will also benefit other used free software projects."

Does the wording need to be a bit more precise? Don't we also want to be able to subsidising/cover the cost of developers? Not necessarily @Moult or @stephen_l but in principle.

Personally from working with the website and forum for a while I've been disappointing to find only the wikilab project as a full project developed (I think @yorik ?) with Free Software and freely documented so we can share it. We really need more examples before I myself even take us seriously. A great website and branding without case stories would be a bit hollow. I think the way Blender made the blender movies is an example of how real world projects can push development. So I'm interested in a building project - even if it is fictitious. All I have managed to scrape together so far is wikilab (https://wiki.osarch.org/index.php?title=Architecture_3D_models_created_in_FreeCAD I hope it gets it's own page) and possibly a shed I haven't quite tracked down yet.

Another avenue to use funding would be administrative funding. Setting up a legal entity (after heavy consultation with similar free software projects) would allow us to provide a legal structure for decision making and seeking funding. If we want to do that I think we should ask relevant organizations for advice on what works best. Who would they be? FSF, Software Conservancy, Linux Foundation, Blender Foundation/Institute, The Document Foundation. I know that might sound like a boring topic but many other things will be much easier when there is an organization behind a funding proposal. Remember that Blender has historically had a lot of public funding, KDE & Libreoffice get a lot of funding from the German state

B by baswein on 14 Aug 2020 #

Congratulations Dion on attracting so much excitement and interest!

As usual I agree with Duncan that a open project is an exciting prospect.

Perhaps a way to achieve this would be a deign competition that requires the design be produced using opensource tools. The building(s) being designed could be for a non profit. The better design competitions that I have been involved in are ones that have a select group of designers who are paid a set amount to produce the design. In this case a RFQ could be sent out that asks people to put together a team of professionals and the opensource pipeline they propose to use. A number of teams would be chosen and funded to some extent to create designs. A winner would be chosen and either given prize money or a contract to bring the building into reality.

M by Moult on 14 Aug 2020 #

@duncan writes: Does the wording need to be a bit more precise?

Unfortunately I haven't yet got any interested non-for-profits where this might actually turn into a reality yet. It was just an idea. Would you like to have a go at rephrasing?

@duncan writes: Don't we also want to be able to subsidising/cover the cost of developers? Not necessarily @Moult or @stephen_l but in principle.

In principle, yes. Developers need to be paid. If I had a big pool of money, I'd immediately give @stephen_l whatever he needs to work full time on what he thinks is best for tools like Archipack. I'd also pay myself at a reasonable hourly rate for whatever hours I put in (I wouldn't trade my day-job, but that's a personal choice). However, as it stands, I don't have a big pool of money, I think @stephen_l has already achieved self-funding and I'm willing to sacrifice being compensated in the short term if whatever funds could go to new contributors to let OSArch grow. Especially contributors with certain programming skillsets which are a rarity in AEC coders. In the future, I think this needs to be a goal: every OSArch contributor should be financially compensated fairly for their time.

Can you suggest a rephrase to the grant application that captures this spirit to become a resource that is able to fund OS AECO development?

@duncan writes: Personally from working with the website and forum for a while I've been disappointing to find only the wikilab project as a full project developed (I think @yorik ?) with Free Software and freely documented so we can share it. We really need more examples before I myself even take us seriously. A great website and branding without case stories would be a bit hollow. I think the way Blender made the blender movies is an example of how real world projects can push development. So I'm interested in a building project - even if it is fictitious.

Yes. I agree. @baswein agrees. When I find the time, if somebody hasn't yet started work on your mad scientist lair project, it is something I'll work on. In the meantime, is there some way you can rephrase the application grant that reinforces the spirit of working on a building project in a team? Perhaps we can propose funding so that a group of 3 dedicated individuals can spend 1 month (fully paid at a rate equivalent to their current employer's monthly rate) to work together full time in producing a building project? This equates to 3 months of salary, which is an amount we can explicitly request. Alternatively, we can propose working half time over 2 months (the half time allows us to make agreements with our current employers, as well as provide a bit of buffer time so that new features can be built whilst we're working)?

What do you reckon, if you think we're on the same page, can you try draft a paragraph that I can amend the application grant with?

D by duncan on 15 Aug 2020, edited 15 Aug 2020 #

Here are some thought made concrete

original:

Partnering with non-for-profits to use free software for designing and constructing buildings for the benefit of society, with funding used to subsidise architect / engineer / building costs. This work will benefit the BlenderBIM Add-on, but will also benefit other used free software projects.

proposed:

Partnering with not-for-profits to use free software for designing and constructing a building or building system for the benefit of society, with funding used to subsidize programmers, consultants and construction. This work will benefit the BlenderBIM Add-on, but will also strengthen the entire software pipeline.

That's my try at a new text.

I've started a new thread for the Mad Scientist project

I by iosvarms on 15 Aug 2020 # +3 votes

@Moult said:

A point of clarification I should mention: the grant does not need to be related to Unreal Engine, Epic Games, or the game industry. If awarded a grant, I personally have no plans to invest it in integrating the BlenderBIM Add-on with gaming and Unreal Engine, as I personally do not see those integrations as currently high priority / high impact to the AEC industry. I think others would agree based on the feature requests seen in these forums. In my discussions with the representative from Epic Games, I have clarified that the BlenderBIM Add-on has little to no relationship to Epic Games directly (I was actually quite surprised to be approached, as I could not see the benefit to Epic Games). Both they and I are also aware that Unreal already supports IFC import. The impression I got from the conversation with them is that this seem to be a non-issue.

@Moult don't be surprised. Epic Games knows there is a market for them in AEC. For example in the Quantity Surveying domain, there is an idea about using 3D-point clouds/VRs/drones/cameras (attached to hardhats) for on-site valuation of works done on site. Also in the Real-Estate sector, virtual tours of properties to be let out is a possibility. The idea of simulating the construction process during planning (Live 4D-BIM). There is also potentials in GIS and BIM. I am certain Epic Games know there will be value if they fund BlenderBIM. Also there are lots of research on this. Check out the Automation in Construction Journal.

https://www.researchgate.net/publication/229339883_Integrating_BIM_and_gaming_for_real-time_interactive_architectural_visualization

https://www.sciencedirect.com/science/article/abs/pii/S0926580519315146

R by ReD_CoDE on 15 Aug 2020 #

@iosvarms there're values, this is why Epic Games support Blender Foundation

And there's a long time Epic Games knows about opportunities in the AEC industry, and even some have built their solutions based on Unreal

But for Epic Games, the place Blender would have and should have is clear, even for Blender Foundation too

The main part that I think you all missed is that first of all you have to know the direction all go, and be ready to be part of that movement

But if you ask me, I will say that you all don't know about these things, but this doesn't mean you can't be ready for it, today or tomorrow

V by vocx on 16 Aug 2020 #

@duncan said:

We really need more examples before I myself even take us seriously.

I think paullee or Bernd mentioned it already; the open source house by Studiolada is something that could be considered a real project that was built; the guys at FreeCAD tried recreating the project; Open Source House with Open Source CAD? This Thomas Neemann guy has been looking into many workflows in which FreeCAD and open source tools could be used for real projects; apparently he has had some success, but there is still more to do, particularly when it comes to handling big files (hundreds of megabytes); Arbeitsablauf für schnelles arbeiten mit großen IFC, STEP, IPT, 3D-DXF, 3D DWG, PLY, ASC, xyz, e57, ptx u. IGES -Files

Another avenue to use funding would be administrative funding. Setting up a legal entity ... would allow us to provide a legal structure for decision making and seeking funding. If we want to do that I think we should ask relevant organizations for advice on what works best. Who would they be? FSF, Software Conservancy, Linux Foundation, Blender Foundation/Institute, The Document Foundation.

Apparently long time ago FreeCAD decided to apply to the Software Conservancy, but the application was never made until recently. Now we are waiting for a reply, which should come before the end of the year hopefully. What this means in funding terms is still uncertain, but hopefully that allows the FreeCAD project to have a better structure, FreeCAD partnership with Software Freedom Conservancy.

D by duncan on 16 Aug 2020 #

@vocx that 'open source' house by Studiolada could be a really good thing too keep working on. I'll look into it a bit more. Thanks.

D by DADA_universe on 16 Aug 2020 #

@duncan said:

Personally from working with the website and forum for a while I've been disappointing to find only the wikilab project as a full project developed (I think @yorik ?) with Free Software and freely documented so we can share it. We really need more examples before I myself even take us seriously. A great website and branding without case stories would be a bit hollow. I think the way Blender made the blender movies is an example of how real world projects can push development. So I'm interested in a building project - even if it is fictitious. All I have managed to scrape together so far is wikilab (https://wiki.osarch.org/index.php?title=Architecture_3D_models_created_in_FreeCAD I hope it gets it's own page) and possibly a shed I haven't quite tracked down yet.

@duncan, I am aware of The Plypad by Marslow CNC (an open source CNC platform) which has open source blueprints available. It's an experimental project which cam be built with the very affordable Open Source Marslow CNC kit. There's also Open Building Institute and Open Source Ecology, both working on building projects within the Open Source paradigm. Marcin from OSE should be making a presentation at the monthly meetup on the 3rd of October.

Y by yorik on 16 Aug 2020 #

Amazing initiative @Moult ! I hope it goes forward.. And I also think there are good chances, blenderbim is well at the crossroads of several things that interest epic games..

As for freecad, still some work to do to build a better structure and attract some bigger $, but we'll get there ;)

D by DADA_universe on 16 Aug 2020 #

@Moult said:

Second part:

Describe your next steps and plan to develop your project

..............................

However, with additional funding that is large enough to hire dedicated developers, I will invest in two known shortcomings as a high priority:

New features to rapidly do architectural and MEP (mechanical, electrical, plumbing) 3D modeling with parametric shapes in a user friendly manner. Blender lacks many CAD transformation features and one-click parametric geometry that is needed to model buildings.

The ability to create 2D documentation (i.e. drawings & schedules) to the standard required by commercial jobs. Current tools are slow, clunky, and not feature complete. Without the ability to create construction documentation to a high standard, no free software will be considered seriously as an authoring platform.

There is no shortage of other things to work on (see commit log), but the two shortcomings identified above are likely to have the highest impact on the AEC industry.

Without enough funding to hire dedicated developers, I am exploring other options to use potential funding, including:

Partnering with non-for-profits to use free software for designing and constructing buildings for the benefit of society, with funding used to subsidise architect / engineer / building costs. This work will benefit the BlenderBIM Add-on, but will also benefit other used free software projects.

Partnering with academics, who are exploring using free software for research projects in the AEC industry. The underlying library used by the BlenderBIM Add-on and FreeCAD, called IfcOpenShell, was started in academia, and academics are usually involved in developing OpenBIM standards which are vital for industry interoperability.

Sponsoring membership to buildingSMART international, the non-for-profit organisation responsible for developing OpenBIM data schemas, including IFC, BCF, MVDs, bsDD, and the OpenCDE spec. This would help give a "voice" to the free and open source software community in developing open standards for the AEC industry. Membership is currently paywalled and given priority to large proprietary vendors.

Not using the funding for the BlenderBIM Add-on, but instead for other free software in the AEC space if we decide that their need is greater than ours. The AEC industry is huge, with much more than one tool involved. This decision can be made with the help of the OSArch community - which encompasses many disciplines and tools in AEC.

Leaving the funding untouched, but in a community fund that may grow with other contributors until we can do something productive with it, or be saved until a rainy day when devs need funding, or until a better opportunity presents itself when the project is more mature in about a year given the current rate of development (e.g. a project wanting to use free software comes along, or an enthusiastic developer appears, etc). I hope that as the project grows, I will, along with the rest of the OSArch community contributors, be equipped financially to ensure that AEC firms can be confident in the tools that we develop. Funding will help gain credibility among AEC firms who are used to proprietary tools and do not take free software options seriously.

No development / spending decisions will be made in isolation. Development is already done transparently based on the many suggestions of the OSArch community - they provide the insights, feature requests, and bug reports of many varied disciplines that need to work together to create the built environment. Any spending decisions will also be made with their consultation and support. This is a large, complex industry issue which is being tackled and needs the community support.

Cheers. Hope it resonates. And if it does, I hope it comes to fruition.

@Moult this is a great development. Please find my feedback below, based on what I've read:

It appears to me that the language of the proposal might need to be a bit more specific. The first two bullet points where you nailed the pain points you would like to prioritize and fix with the help of the grant were very specific. The other list of things you might consider spending the funds on if the grant was not enough to engage developers full time seem better suited for internal debate here. What should eventually make it into the proposal (I think) should be more specific. While in agreement with the need to build / support the larger OSArch ecosystem, it seems counter intuitive to plan to do this with a smaller pool of funds, supporting the larger ecosystem would most definitely cost more than polishing BlenderBIM to a point where it can compete favourably with ''Industry Standard" tools. Especially since getting BlenderBIM to that point would already be supporting the larger ecosystem in many ways.

"Not using the funding for the BlenderBIM Add-on, but instead for other free software in the AEC space if we decide that their need is greater than ours." Makes it read like BlenderBIM does not really need funding. Then why apply? Shouldn't the other free software with greater needs apply instead? I think that sentiment confuses the messaging and might not give confidence to sponsors that the project is laser focused. Ditto leaving the funds untouched till a rainy day, or till we're in a better position to use it....if it's not a rainy day, or the capacity to utilize the funds is not there, then why apply? No corporation will honour a grant application where there's no proven need nor capacity to utilize intended funds.

D by DADA_universe on 16 Aug 2020 # +3 votes

Looking at previous Epic Mega Grants, my observation is that they fund tools, yes the communities behind the tools benefit from the grants, but that's on the basis of, and as a consequence of the tools being developed in the first place. I therefore suggest that the focus / priorities for the grant application should be staged as follows:

Hiring dedicated developers to focus on delivering top notch new features, CAD functionalities, 2d documentation to BlenderBim.
Identify a set of tools already in Blender you'd like to extend their integration with BlenderBIM even further towards achieving #1. You are already collaborating with Archipack, MeasureItArch and others. This work can be deepened and fast tracked with funding.
Work on interoperability with other tools in the FLOSS ecosystem for the AEC industry, as being done with FreeCAD, etc. No matter the size of the grant that eventually comes through, I think these three should be the primary areas of priority, and you can scale up or scale down the operation / expenditure to fit the sort of cash flow the total grant amount would accommodate accordingly.

If these are adequately catered for, then secondary areas of focus can be attended to, this being without prejudice to their importance in the overall scheme of things, but with an eye on strategic phasing of activities. For example, sponsoring members to BuildingSMART would get us a lot more mileage when we have more polished tools out there and have gained a lot more adoption in the professional space with several use cases and live project scenarios to point at. So tools first, then -> community (of users and professionals) and interoperability with other FOSS tools in the space, then -> the larger AEC industry, generally in that order, even if not strictly so. Plus, please be bold and specific, it's a great thing you're building and it's got good legs.

V by vocx on 16 Aug 2020, edited 16 Aug 2020 # +1 votes

@Moult said:

I have recently been contacted by someone from Epic Games, related to Unreal Engine. One of the things we discussed was the Epic Games Megagrant and whether the BlenderBIM add-on should apply. I have drafted up an application below. I plan to submit early next week.

It is surprising that you were contacted, and not the other way around. In FreeCAD we became aware of this grant in February, About funding and all that. The way Epic describes it, the grant is very generous because they are willing to give money without really expecting anything in return, particularly if the amount requested is less than USD 25 000. For amounts bigger than that, they do expect some sort of communication and updates on progress.

@DADA_universe said:

It appears to me that the language of the proposal might need to be a bit more specific.

I agree with @DADA_universe that you need a real course of action, and not just a description of what you want to accomplish. Your "second part" reads more like a list of wishes and not like actual steps.

The wording of the grant encourages single individuals to apply, not necessarily big organizations. With this is mind, I actually applied for that threshold of USD 25 K, and justified it by describing what I would do with the money.

2000 to get new computing equipment, CPU, monitors, GPU. The rest for the salary of a full time programmer.

And I described the activities that I would do within FreeCAD, which I am carrying anyway in a voluntary fashion.

Add proper unit tests to the Architecture and BIM workbenches.

Split commands of the Arch Workbench and BIM into a programming interface and graphical code (buttons), similar to Draft.

Improve the IFC importer (single core and multicore).

Continue Draft refactoring in terms of Python style, and docstring documentation.

Continue Draft refactoring of code in new modules; import and export of SVG, DWG, DXF; Vector utilities, geometrical utilities, working plane.

Add support in Draft for libredwg (DWG and DXF import library).

Update the generation of the Doxygen documentation, using a newer Doxygen configuration file; use Sphinx for documentation generation of Python; use doxypypy to pre-process Python for use with Doxygen; use Breathe to pre-process C++ code for use with Sphinx.

Improve the Addon Manager to make it a library, so that it can be used without the graphical interface being loaded.

Improve the Workbench-Starterkit to make it easy to create new workbenches.

Migrate external workbenches to use the new-style workbench structure of "namespace workbenches".

Improve the FreeCAD tools to generate the offline documentation from the wiki.

Improve several tutorials in the online wiki.

Improve the example workbenches and scripted objects in the source code.

Improve the power users and programmers documentation in the wiki, particularly as it relates to creating new scripted objects, new graphical commands, and new workbenches.

Improve the Draft Grid so it is more like Blender's.

Improve Raspberry Pi 4 compiling and testing.

I intentionally added a lot of topics to show that there is opportunities to improve FreeCAD in many ways, and I have contributed in all those areas a bit already, and a ton to the FreeCAD wiki. Considering my own rate of development, I feel this would take me more than a year to accomplish, so I also mentioned that if they were willing to fund double, that is, $50 K, I could hire a second developer to advance these topics faster.

I was a bit optimistic that they would answer quickly, as back then in February I had already committed many things to FreeCAD, had refactored a lot of Draft and Arch code, and more pull requests were on the queue. Unfortunately, I made my application in the middle of March, when the pandemic hysteria was at its highest; by that I mean locking down absolutely everything and hoarding toilet paper; at that time it was clear that many companies needed to save as much cash as possible to ride the wave. So, I received a reply three months later, in June (they mention they give a reply within 90 days), not straight up rejecting my application but saying something like "We need more time to review your proposal". To me this means it won't be successful after all.

Separately from that, I was recently hired (and then fired) by a company that wanted to improve the IFC importer in order to do some energy calculations in buildings; but the funding was cancelled so the project didn't continue after one month.

So, my conclusion is that modelling tools like FreeCAD and Blender are on the brink of major breakthroughs. I think Blender is a winner already, and FreeCAD needs a bit more organization before it can cause a big impact in different industries. My personal advice is to take the opportunity of getting funding; even if the next steps are not entirely clear, trying new things, and moving forward is clearly the way to go. However, I do like that you keep the commercial focus of Blender-BIM, so that it isn't just some experimental project with no clear objective. If you need any help, let me know. I haven't really used Blender-BIM because I've been heavily invested in FreeCAD, but maybe I can also shift my focus to Blender, more so if there is some coin to be made.

M by Moult on 26 Aug 2020, edited 27 Aug 2020 # +11 votes

Sorry for the radio silence - I've tried to take everybody's views into consideration and have rewritten the application grant as shown below.

Things I have considered:

We all agree that ideally, time spent on a real building project is the best way to go. However, we don't have one right now, and I do not want to forego potential funding simply because the most ideal project isn't currently available (I have reached out to various non-profits without success). There are other priorities we can tackle in the meantime, and I've tried to choose ones that will place the tools with a very strong foundation, when the real building projects start coming in the future.
Hiring dedicated developers, upon reflection, is complex, due to the large domain understanding required for IFC. I expect that investing in core developers will also provide a lot of value.
I have reworded to be more precise on exactly three tasks which I have identified to bring as much value to OSArch, BlenderBIM Add-on, but also FreeCAD, BIMSurfer, BIMServer, and as many FOSS projects as possible. You will note that these tasks do not necessarily provide immediate benefit - this is because things that provide immediate benefit are likely to be implemented anyway in volunteer time. However, these things are a fundamental investment into the future of OSArch and its umbrella of applications.
I have placed an emphasis on the sustainable financial resources of OSArch

Describe your next steps and plan to develop your project

There is no shortage of other things to work on (see commit log), but I have identified three shortcomings that when solved are likely to have the highest impact on the AEC industry, and without dedicated attention would likely take a very long time to be resolved. The three are: reliability, drawing generation, and data specification compliance.

Reliability: Geometry in IFC is procedural and complex, therefore error-prone. Geometry issues can lead to wrong quantities and distrust in models and project partners. We will build a comprehensive test suite for IfcOpenShell and OpenCASCADE to ensure stable geometry processing for all dependent FOSS projects. 6000USD

Under the hood of the BlenderBIM Add-on, FreeCAD, BIMSurfer, and many other free software that builds technology for BIM is a reliance on a library called IfcOpenShell based on OpenCASCADE, which does geometry processing. This geometry processing is highly fundamental and complex, especially given the diversity of datasets in the industry. To ensure that a stable foundation exists for all AEC FOSS projects in the future that deal with IFC BIM data, we need to complete building a test suite with a continuous integration server for different build environments, along with a way to continuously check geometry processing errors.

Without this investment, in the future we are likely to run into stability problems that shake the reliability of many AEC FOSS projects simultaneously.

Work has already started by Johan, one of the main IfcOpenShell developers, but there are hardware costs involved, as well as man-hour costs required to finish the job. Tasks include:

Result browsing and filtering, graph plotting

Visualization in web viewer

Metric calculation, e.g. group types of errors

Extract details from IfcConvert log files

Web-based presentation e.g CSS files

Interactive and continuous processing of new commits

Drawing generation: The AEC industry operates largely on sectional drawings and symbology for execution and communication because the notation enables experts to quickly consume the information. We will re-engineer the C++ IFC to SVG serialiser for construction drawing generation so that it can be accessed via Python, adding multithreaded support, improved caching, standardised interfaces, and optimised data handling. 4000USD

A fundamental feature in AEC is the conversion of 3D geometry to 2D construction drawings. Various approaches have already been attempted that do this conversion to varying quality. To finalise this feature requires consolidating these approaches into a C++ library that focuses just on this task. This will form the foundation of all drawing generation in the future, across multiple FOSS projects.

This task is difficult as it requires specific domain knowledge of IFC geometry and an understanding of the OpenCASCADE geometry kernel with its many quirks and workarounds. Costs will sponsor man-hours directly.

Specification compliance: reliable data exchange between disciplines / fields in the AEC rely on compliance to OpenBIM and ISO open data format specifications. We will add support for at least 80% of the IFC specification via the BlenderBIM Add-on. 5000USD

The BlenderBIM Add-on currently supports 69% of the concepts described by the IFC specification. In order for Blender to be useful as a tool not just by architects, but also for mechanical, electrical, fire, engineering, environmental analysis, and structural analysis, it needs to be able to consume the metadata required in the lesser known aspects of the IFC specification.

Time invested specifically in implementing the specification will mean that in the future, Blender will have the fundamentals already implemented when more non-architectural disciplines want to use free software, and we can focus instead on workflows and polishing the user experience. This investment will dramatically increase the potential and usefulness of Blender as a tool in AEC. Costs will support man-hours.

These three initiatives (totalling 15,000USD) will be primarily done by three core developers (Johan, Thomas, and Dion (myself) respectively) - who will focus on our respective areas of expertise. Throughout this process, we will also engage with the open source AEC community (OSArch.org) to ensure that we are on track.

The successful receiving of funding will also help achieve a goal of the OSArch community - where we hope that FOSS tools in AEC can be sustainably resourced, and not purely depend on volunteers. It will also help signal to the currently proprietary market that FOSS in AEC is a growing field, and support future partnerships with AEC firms.

This application grant was done in a transparent manner, in consultation with the OSArch community: https://community.osarch.org/discussion/165/application-to-epic-games-megagrants-for-the-blenderbim-add-on

J by JQL on 30 Aug 2020, edited 30 Aug 2020 #

Hi to everyone but especially @Moult ,

I've read this with interest but the final application text seems a bit more criptic to me than the first.

I'm saying this in a "potential user" perspective and probably I shouldn't say any of it as I've not had the chance to get involved with BlenderBIM or this community that much. On the first version of the text, the two bullet points were clearly stated and were clear needs for the end user:

Features for MEP
Standard 2D Documentation

Now it looks like you are not going to address the final needs themselves for the user, but will address core functional aspects instead, that they will be built upon. I would say that the strategy of building the foundation as solid as possible is the best way to go. This has the added plus that the foundation is common to all that free software, which is excellent for everyone. However, I think it would be worth it to make the application for the Epic Games Megagrant to extend to the specific development of the BlenderBIM addon in order to reach the end user. Without a user base, BlenderBIM won't grow and we (or I), as potential users, are not yet fully convinced that we can be productive with it.

People see huge potential in BlenderBIM on two aspects: OpenBIM and Blender as a foundation. Blender has proved its strength on the 3D world and it's credibility for archviz. we know we can model everything with it, but it stops there. We are not modellers we are architects and we need those features.

However, Blender is solid. It is a great start. What you're doing is also solid and has a clear future. Honestly, I don't know if all other software in the OSarch started as strong as this, and even if they have done more and are at more advanced states, maybe that's the reason Epic Games addressed you.

So, I would think that building the foundation would be a first stage of the grant and then development of those specific Blender BIM needs the second stage:

Stage 1 - Core functionality shared with OSarch software.

Reliability;
Drawing generation;
Specification compliance.

Stage 2 - Main BlenderBIM functionality

Features for MEP in BlenderBIM;
Standard 2D Documentation Tools in BlenderBIM.
DWG/DXF export instead of SVG - I took the liberty of adding this one as it's mandatory to deal with consultants and going through inkscape is not viable imho (There might be other paths to DWG/DXF that I am not aware of besides Inkscape. I didn't find them so I stopped thinking on BlenderBIM. I'm ashamed but it was as simple as that.)

What I also think is that Blender is as attractive as it is scary, and as a potential user I would also add the following stage of development in order for BlenderBIM to be able to grow:

Stage 3 - Full scope tutorials and documentation

Exhaustive documentation and Video Tutorials for BlenderBIM with one or more recommended workflows.
Documentation and Video Tutorials on the best modelling techniques for architecture in Blender that are compliant with BlenderBIM workflows.

This can probably be built by non coders in the community here, but if it is not consistently addressed, do you think you'll have access to the users that will help you build strength in BlenderBIM?

I'm willing to try it, but I haven't seen either a tutorial that goes all the way through a project or basic project modelling, management of data and documentation, I don't know how to think in BlenderBIM terms, or what sort of things I should start with or what kind of modelling organization should I create. I'm not willing to spend several weeks exploring Blender and BlenderBIM, risking getting to a dead end. I have not yet found proof that I will get through a building project if I take the journey. Even the journey is hard to know as simple things as exporting to DWG/DXF don't seem that simple to me. You should also consider people that might be transitioning from CAD to 3D modelling and/or BIM, because many Architects are still drawing in 2D, they're not 3D modelling or using 3D at all. Maybe they will only transition from 2D if they get into BIM. These Architects have options in the market that they know they can trust and BlenderBIM is not one of them yet.

If you solve this latest stage, maybe the real building example will solve itself by real users.

I hope this post isn't offending anyhow as it may only reflect the amount of information I could gather about BlenderBIM, and therefore, my ignorance on the subject. It may be missing some facts about BlenderBIM that I am not aware of. This ignorance is probably what should be solved for the project to thrive but we users are probably both lazy and have only so much free time. Maybe what we need info that is right in front of us. I just hope my post helps somehow.

D by duncan on 30 Aug 2020, edited 30 Aug 2020 #

I'm willing to try it, but I haven't seen either a tutorial that goes all the way through a project or basic project modelling, management of data and documentation, I don't know how to think in BlenderBIM terms, or what sort of things I should start with or what kind of modelling organization should I create. I'm not willing to spend several weeks exploring Blender and BlenderBIM, risking getting to a dead end. I have not yet found proof that I will get through a building project if I take the journey. Even the journey is hard to know as simple things as exporting to DWG/DXF don't seem that simple to me. You should also consider people that might be transitioning from CAD to 3D modelling and/or BIM, because many Architects are still drawing in 2D, they're not 3D modelling or using 3D at all. Maybe they will only transition from 2D if they get into BIM. These Architects have options in the market that they know they can trust and BlenderBIM is not one of them yet.

Totally with you there in your picture of the end user. You are exactly where I am and we have some work ahead of us in finding and documenting solutions. You can see progress on that in some of the other threads and on the wiki.

I'm not totally with you on thinking that the application should focus on that for now. We have some fundamentals that need work before there is an end user friendly product to package.

M by Moult on 31 Aug 2020 # +2 votes

@JQL - welcome to OSArch! Very happy to read your message, regardless of whether or not you've had the chance to play with free software options yet. I'll try and address your concerns in detail - but for the record, the application has been submitted now. There is a 90 day waiting period for a response. I hope the response below clarifies the logic.

@JQL posted:

Now it looks like you are not going to address the final needs themselves for the user, but will address core functional aspects instead, that they will be built upon.

Correct. The logic behind this is that in the everyday development of the BlenderBIM Add-on, the priority is on user needs. Therefore, even without a grant, user needs will be addressed. The grant provides incentive of tackling some of the harder, more fundamental, less "sexy" / visible aspects of the system.

@JQL posted:

Features for MEP in BlenderBIM;

It is not possible to implement this without having specification compliance for the MEP aspects of IFC. This is part of the "specification compliance" portion of the application grant.

@JQL posted:

Standard 2D Documentation Tools in BlenderBIM.

This is a super high priority to solve. It is also a huge beast to tackle. The good news is that it is progressively being solved - nearly every release has some new features and fixes related to this topic, and some releases focus on it completely - for example the last release, where 41 out of 70 improvements were related to solving this. This issue will not be ignored - the application grant focuses on a foundational aspect of it, but in the mean time, I expect to be dedicating a lot of attention to this problem nevertheless.

@JQL posted:

DWG/DXF export instead of SVG - I took the liberty of adding this one as it's mandatory to deal with consultants and going through inkscape is not viable imho (There might be other paths to DWG/DXF that I am not aware of besides Inkscape. I didn't find them so I stopped thinking on BlenderBIM. I'm ashamed but it was as simple as that.)

This is complicated to give a short answer. The output format is heavily related to the drawing generation process. There are two problems here: 1) calculating vector output, and 2) dumping that vector data into a particular format. The latter is easy to solve (for DXF, not DWG). The former is not. In fact, here's a new feature that lets you output in DXF: https://github.com/IfcOpenShell/IfcOpenShell/commit/6f6a1aede64c51d90c4e572baa55221b2ef56a3f - and below is a screenshot of example DXF output:

I think we are aligned in terms of our needs and wants - just that some things from a technical standpoint are quite tricky - we're doing the best we can - and we aren't alone: there are others trying to solve the same problem in the proprietary world, and they haven't cracked it yet either. There's a reason why almost no proprietary CAD app promotes the ability to produce documentation output directly from IFC.

Sorry for the rather short answer. Happy to elaborate. Perhaps best done through voice at the next monthly meetup.

@JQL posted:

I'm willing to try it, but I haven't seen either a tutorial that goes all the way through a project or basic project modelling, management of data and documentation, I don't know how to think in BlenderBIM terms, or what sort of things I should start with or what kind of modelling organization should I create. I'm not willing to spend several weeks exploring Blender and BlenderBIM, risking getting to a dead end. I have not yet found proof that I will get through a building project if I take the journey. Even the journey is hard to know as simple things as exporting to DWG/DXF don't seem that simple to me.

Your impressions, though unfortunate, are accurate. The easiest answer for this is that the BlenderBIM Add-on is not yet ready to deliver a building project, therefore, there are no tutorials for this because it cannot be done (unless you are a crazy super-user). See the roadmap - we are in the "version 0.0.X series" phase. The software, by definition, is incomplete. The BlenderBIM Add-on website provides a warning on the homepage and the download page. Certainly, there are pockets of it which are more mature than others and can be used for isolated tasks in project delivery in a commercial setting, and I use it daily at my work on large complex projects. However, these are only pockets. I have used it to deliver a project, but I use a lot of workarounds and the projects I have used it on are very particular / bespoke. It was only in the last two weeks that you could add a door with a door swing in one click.

If incomplete, alpha software doesn't frighten you, please have a play with the BlenderBIM Add-on, and please ask many questions and request features - even simple ones that we take for granted in mature proprietary software. My YouTube channel has some (forever outdated, due to the pace of development) examples.

R by ReD_CoDE on 31 Aug 2020 #

I forgot to mention that the Epic Games fund is for "registered" open source organizations