OSArch-aeology! ifc and BlenderBIM for excavation documentation? Feedback desired...

Dear OSArch community,
I could really use your input on some ideas, that some of you will hopefully find intriguing... Planning to write my master thesis in archaeology on open source 3d documentation methods for excavations, inspired by other Blender-uses, I came across BIM, and BlenderBIM.
For the past few days, I have now been learning about BIM and ifc (basically just grasped the basic concepts), to see whats possible. As archaeological excavations are still largely documented in 2D, and the 3D models we create of our trenches and layers (photogrammetry or laserscans) are usually treated as extras, rarely connected to data, I see huge potential for improvement in archaeological practice.
IFC got me really interested because of the way it structures data.
Regarding spatial structures and aggregates: Just like in AEC a project contains a site contains a building contains a storey contains a wall, so does in archaeology a project contain a site contains an excavation area contains a trench contains layer contains objects.
Regarding materials: Just like in AEC a wall can consist of multiple elements, an archaeological layer can consist of multiple elements. And just like another type of IFC material classification (constituent or smth?) can describe cement as consisting of water, sand etc., we describe layers or encountered architectural remains as consisting out of this or that material in this or that amount.
Other super useful and important features for then 3D modeled excavation sites would be geolocation, the sectioning at any desired spot and the 2D Plan generation offered by BlenderBIM.

So I hope you see how I consider the general way IFC structure data as very similar to how we structure data in archaeology, although we often do so on paper and in spreadsheets...

Now, regarding attributes and property sets, this system would also match well with archaeology, as the attributes of many IFC classes (that cannot be changed) are desirable as they are for archaeological features, and custom property sets would allow us to set all those archaeology-specific properties, like dating of a layer, subjective description, samples taken, photographs linked etc. (I think?)

I am aware that custom properties of course partly removes the "structured-ness" and standardisation from the data, which is the point of IFC, but in archaeology some people may actually prefer this customizability of the properties over standardized input due to the huge variation in archaeological features around the world.

So instead of our current relational databases based just on words and numbers, with 3D models linked like photos, used just as extra-data one can look at if desired, I imagine a 3D model of a trench with all its layers and objects that IS the database, just like BIM, but without the "Building". So this would be very different from the uses of BIM in the heritage sector so far (like HBIM and ArchaeoBIM), that all still focus only on architecture.
(And there are good workflows recently created that end up with 3D models of the individual layers with a reasonable poly-count, so that aspect should hopefully not be a problem)

As you can probably gather, my learning-journey on this topic just began, and my knowledge is quite superficial thus far. But before I get into this too deeply, I wanted to ask for your perspectives on this idea.
Do you think this would work? Or should we archaeologists stick to our QGIS polygons with their attribute tables and such?
If you think it makes sense from the basic structure, do you think there are IFC classes/concepts that could serve for this purpose? Or by misappropriating classes for this purpose could I at most make a proof-of-concept for structuring data in this way, but we would need very different classes for archaeology?

Would hugely appreciate your insights on this. Soon I am set to present my ideas on this to my fellow archaeologists, who do not know about BIM and certainly not about IFC, so your knowledge is highly relevant. My master thesis (as currently planned) will not solely revolve around this, but possibilities of open source 3d documentation of excavations in general, so feel free to very much torpedo my hopes for this :D

Thank you :)

CoenbasweinAcebitacovirwmi

Comments

  • Do archeologists 3D scan point clouds?

    Maybe something like this could help your thesis?

  • Yes, either with Terrestrial Laserscanners or calculated through photogrammetry we document layers and other features we encounter. Making 3D models out of those in Blender is not a problem and has been done before. What would be new about my idea is turning these models into data-containers for all excavation data, using ifc if possible. That is what I would like your opinions on.
    But thank you for that first pointer.
    Another workflow that people tried in recent years is importing these 3d models into ArcGIS and attaching attribute tables to them there, basically doing something similar. But I still see many advantages if we could use openBIM Methods for this, with BlenderBIM directly in Blender.

  • This might be a good idea. If there is no other existing data format that allows you to attach archaeological properties, external links, and data to 3d data then IFC would be able to do it. It is also has as good a chance as anything else of being readable into the future.

    A next step would be to try a simple test model to see what can be achieved with the existing tools, this will reveal what new tooling and standards need to created.

  • @Coen said:
    Do archeologists 3D scan point clouds?

    Maybe something like this could help your thesis?

    I think point clouds can be stored in an ifc, so what Coen is getting at is that this can be the basis for your model/data driven approach to documentation.
    Your closest field mates are the heritage and conservation people who deal with similar techniques I think we've had a few good discussions around this:
    https://community.osarch.org/discussion/comment/8022/#Comment_8022
    this conversation, and this one:
    https://community.osarch.org/discussion/comment/12967/#Comment_12967

    You can ofcourse save custom meshes and assign info to it that was, BlenderBim is very flexible in that way! and I do this quite a bit, the model isn't as efficient though.

    It would be very cool to see a proof of concept! Nothing you've described is not feasible with Ifc and BlenderBim.
    It would also be nice to add the A of Archeology to the set, a common joke at the beginning of every construction project that we hope we find no bones, because then we would need to report it and get shut down by the archeology people hahahha

  • @brunopostle
    Thank you. We have no data formats dedicated to only archaeology, so we make do with GIS, which is of course very handy for different spatial analyses relevant to archaeological questions, and the 3d capabilities of ArcGIS are starting to be looked at more closely by archaeologist in (very) recent years. Turning to open source GIS however, the options on what is possible with 3d data is much more limited, but slowly growing. IFC in my eyes (and hopes) could maybe be an alternative to 3d GIS, or maybe complementary, and since it can be plain text if you want it to, I considered it quite future-proof. Any opinions on that are very very welcome (also if just reinforcing previously voiced ones).

    Yes, a test model would be my next step after getting some opinions. Guess it is time for me to dive deeper... maybe IfcExternalSpatialElement could be a class for archaeological layers?

    @Ace
    Haha yes, indeed, apologies on behalf of my guild for any scheduling we mess up. And as a side note, if it did work out I played with the idea of calling the approach AIM - Archaeology Information Modeling, to distinguish it from building stuff, but not sure if that is what you meant...

    Hm, I will check if it is less work (or has other advantages) to start with a point cloud in ifc, rather than starting with the custom mesh modeled from the point cloud (or modeled from the total station survey).

    And thank you for the discussion links, I only came across the second one when I searched the forum for "heritage" and "archaeology" etc.
    I did read some articles on Heritage BIM, and their greatest impediment is creating all the custom building elements and such, while I believe my greatest hurdle is still to figure out which classes I would have to assign to the archaeological features to conform to IFC hierarchies and categories. But there will probably be more hurdles to overcome once I learn more.

  • I encourage you to explore the potential of BIM-IFC in your workflow. Even if at the end it is not fully adaptable. Different lines of actions that I see:

    • You could use BlenderBIM for your workflow.
    • If you find that you need costumed features, you could create an addon to integrate into BlenderBIM
    • If you find that an add-on is not enough, you could fork BlenderBIM to hack it and create your own BIM app for your job (and create your own community, as well).
      And you can... because all here is FLOSSSSSSS...
  • @ktm said:
    I did read some articles on Heritage BIM, and their greatest impediment is creating all the custom building elements and such, while I believe my greatest hurdle is still to figure out which classes I would have to assign to the archaeological features to conform to IFC hierarchies and categories.

    Probably you would make everything IfcGeographicElement, then create a series of USERDEFINED ObjectType sub-classes to cover your needs. You can reassign classes later if necessary, but this ability to create sub-classes is very powerful and should work fine.

  • Thank you, I will look into that.
    If I get to do a proof-of-concept of this for my thesis I will let you guys know.

    Ace
Sign In or Register to comment.