Different insert points for branches

Hello once again :slight_smile:

I have now tested Speckle a bit, made an initial pitch to my professor who supervises my master´s project - and we came to an idea to potentially integrate Speckle in a “digital twin” course (still very very early in a testing phase without a safe commit to later using it though, to be fair).

So the basic setup would be: Every student builds their house models in revit on an individual “plot”. Currently, the professor would load every revit model / rvt project file into a “main” project at the end of the semester by hand, resulting in a “digital twin city”.
Afaik, you can have one “main” rvt project file in speckle, which you publish as the main branch of a speckle project. Further, you can send other rvt project files (aka the students houses as a specific branch to the speckle project. The main rvt project could then receive the branches as “individual houses”. Thats at least what I tested.

When I set the reference point to of the branch model to be send to the projects default/internal origin, the main branch receives the model according to its internal origin - if the branch model is e.g. on the top-right of its own internal origin, the main branch receives it at its own top-right. So far so good. Theoretically, one could then adjust each branch model relative to their origin, so they don’t overlap.

Question: Is there a reasonable way to define for each branch in the main branch / main project, where it should be placed? So e.g. I set a point A in the main model, assign this point as the insert point for the branch A, resulting in the branch A´s origin (aka the students model) coincides with the “insert” point A in the main model/project?

Is the matter understandable, or a bit too abstract? :slight_smile:

Thank you very much

With best regards,

1 Like

Hi @maniwo! The way I’d articulate this is by having a shared masterplan of “plots”. Each student could receive this “masterplan” and thereafter design within their assigned boundary.

So a hypothetical structure could be:

By having a shared masterplan (even a simple grid?) you remove the need for complex reference point exercises.

Does this makes sense? It’s a bit of a more abstract process that can work with multiple software too - for example, the masterplan could be coming out of Civil3D or Rhino.


Hi @maniwo,

To add to @dimitrie’s answer (with which I HARD AGREE), I can’t stress enough how important it is always to consider how to coordinate Revit with non-Revit software in mind, whether you’re using Speckle or not.

Simplifying their process might be ideal, depending on what you’re aiming for your students to learn. If mastering Revit is a part of your course objectives (brace yourself! :sweat_smile:), I have more potential suggestions that could be beneficial.

I’ve many years (and battle scars) of handling situations like this pre-Speckle, so I’d be happy to discuss further strategies with you in a call if you’re interested. It’s always exciting to see how we can improve the learning process! :mortar_board::rocket:

1 Like

Hi @dimitrie .
I´ve thought about such a “simple” outline in the first way, and would have tried it this way.
Our Prof would have made such a plan after everyone had built their houses (see plan and “result” of last semester below), but making the layout beforehand would be actually less workload.


I just thought about the following: Until now, every student got their 20x20m base plot (by means of property lines, terrain, and a few trees :slight_smile: ). If they - by accident - would move their plot in their own revit project i.e. sideways, and send their models - would this potentially lead to clashes? Well, kinda redundant question if everyone actually has the “whole” model / city plan… Just to explain, why I have asked that question…

But to make sure, if I understood it correctly: The students could either receive the whole Masterplan, with all - say - 30 plots; or the owner could send the plots by selection to a branch for each student - if the student would send their buildings, they get placed where the plot is located?

Or do I misconception an important part? :slight_smile:

Secondly, I just tested a rough “workflow” of this with a second computer - the second has the masterplan where all the models/branches are received, and the first had a single plot (to make it an easy exercise). I send from the first computer a test model (simple structure of walls, floors, stairs, revit objects [windows, doors, …], and non-revit BIM objects). In Speckle everything looked fine, even in Blender. It also looked fine If I received the model in another revit project on the same computer.
However, on the second PC, it imported the non-revit models as either generic cabinets (for furniture) or a generic toilet (well… for a manufacturers toilet - also it lost any other information, which would have been send/displayed on the same computer). I have tested it both with send/receive linked models on respectively off on both computers - am I missing a step there? (I have send “everything”)

Note, that it also displayed an error, that some of the models could not be placed due to them could not be rotated.

Thank you once again :slight_smile:

Both of these would end with the same outcome (assuming they don’t move their plots)

I’m not 100% what you are saying, but in principle a Revit object sent and received in another software and edited is no longer a Revit object. It is indeed lossy.

What you might be looking for is functionality available within the Mapping tool, but this is currently only available for some element types in Rhino (soon Sketchup)

Hi @jonathon .
Well first of all, I´m not exactly the dude teaching them how to use revit :). But I have some “stakes” at the course :slight_smile: . We´re further not exactly an architecture, construction/civil engineering, … degree - we view it more from the perspective of managing data (but by any accident, our Prof is a construction engineer, so one thing led to another).
Tl;dr: Mastering Revit is really not the main objective of the course, but more an incentive to some :slight_smile:

Hm ok - I´ll try to explain/recap it differently - so sending/receiving virtually from Revit on PC1 to Revit on PC2 (by Speckles conversion ofc), resulted in my case in geometries (placed BIM Models by default NOT included in Revit) converted to a “similar” generic model (lights to generic cabinet, furniture to generic cabinet, …). However, viewing this model in the Speckle Viewer prior to receiving it in Revit on PC2, it displayed the correct geometries (and information)

Sending/receiving from Revit on PC1 to Revit on PC1 (through a branch; different projects) contains all the right models and information.