Could SketchUp be transformed to a BIM or PEN System?
-
@honoluludesktop said:
As far as I have experienced, a program that is capable of doing everything compromises its parts.
Please understand that I am not talking about a program but a scheme, something more akin to the characteristics of the web. People use specialist devices to participate in the process of creating or even using a building. Models - components, sub assemblies, buildings - attract and hold data and interpret it for various needs. There is no passing of output. The only physical output is the real thing.
I think, if I might say so, you see computer programs as solutions.
They are not, but they could become enablers once the method of the working was changed. This would define the real objective - for Architects I hope this is the creation of a building not a beautiful set of drawings - and open up use of the web with multimedia to communicate the instructions that get buildings designed and built.
-
The good thing about SU is the way we can sort of directly manipulate the model by pushing in and moving right until we are satisfied. Using BIM packages still resembles more filling tax forms than sketchupping. That's why I am still dreaming of a SU with a more direct link into BIM rather than increasing the BIM'ishness of SU.
Anssi
-
Anssi, I agree. Its like a "one-in-all" solution that takes 10 steps to dimension a edge, vs. a dimensioner that will do it in one step. If however, someone comes up with a "all-in-one" that is all in "one step", then OK, I will be among the first to stand in line to get one (that is, after the first service release:-).
Chris, what you propose is conceptually elegant, but beyond my ability to grasp "the mechanics of". Besides, I like beautiful "working drawings", and will be sad to see the day they are gone, just like my overnight color pencil sketches.
Sorry for all the "off posting".
-
@honoluludesktop said:
is conceptually elegant, but beyond my ability to grasp "the mechanics of". Besides, I like beautiful "working drawings", and will be sad to see the day they are gone, just like my overnight color pencil sketches.
hi all, for my two pence worth,
I also, I like my working drawings, and I want to bury all the boring details inside them.
I can do this by adding layer upon layer of digital information, but I need a digital management tool to navigate quickly from grouping to grouping.I see Chris's tool as a PushPull tool for digital information, which lets me stay in SU longer, giving me more time to do additional renderings, which being part of the single information package can be presented at every possible opportunity I choose. Each project contains all it's own ideas, content, ingredients, I find that fantastic.
I'm trying to help get this working on Macs to give me the single document functionality that is already available to the brave PC user's amongst us. No doubt it's got a way to go before it's a day in day out functional tool, but it can be if people help develop it.
john
keep up the good work chris..
-
@anssi said:
The good thing about SU is the way we can sort of directly manipulate the model by pushing in and moving right until we are satisfied. Using BIM packages still resembles more filling tax forms than sketchupping. That's why I am still dreaming of a SU with a more direct link into BIM rather than increasing the BIM'ishness of SU.
Anssi
Yes that is exactly why I was attracted by Sketchup. I just want to make it better by capturing information at the very moment it's needed to make a model. In this way you have both the model and its data available for scrutiny all the time. If you want to delay a decision about placement or procurement you can test the financial and temporal consequences prior to doing so - perhaps a basis to detach the building industry from its outmoded ties.
@honoluludesktop said:
Chris, what you propose is conceptually elegant, but beyond my ability to grasp "the mechanics of".
The mechanism is the concept. I am sorry I find your post surprisingly paradoxical and reactionary, so cannot really comment further.
@driven said:
I see Chris's tool as a PushPull tool for digital information.
Yes I think that it is a good way of putting it. The key, like bar coding, is to give components identities and link them to a digital record. Given today's amazing communications technology data can be pulled and pushed around to provide pertinent multimedia displays.
@unknownuser said:
Each project contains all it's own ideas, content, ingredients, I find that fantastic.
Not just the project but also each component.
Thanks for your support and particularly your efforts to make it Mac compatible.
-
@mike lucey said:
I am of the 'old school', I started with a pencil, drawing board, T-Square, Set Square and Scale Rule and later on an adjustable? ...... God I've forgotten what it was called ..... eeeeerrrrr .... hold on! I have it now, Adjustable Set Square! Boy were they handy, saved all that messing around with the protractor
I had no difficulties in producing plans, elevations, sections and isometrics also the occasional perspective view for jobs. When computing came into the picture it should have made things simpler BUT this was not the case IMHO. Things became somewhat more complicated for me until SketchUp came along. I initially worked on MicroCAD, this was a decent attempt at 3D. It ran into financial difficulties and was shelved. I was then stuck with AutoCAD ..... the adapted Word Processor Program for draughting!
Mike
Very good comment you are making. I think what made these "old School" production methods work so well is that you were intimately aware of what you were doing. With all this technology we have lost awareness of what we are actually doing. I'm looking in to making a ruby script that can make the classic architectural "axon" using a 45/45 triangle. Simple to do on paper - once you know how to do it - and, I hope, simple to learn to script. Programming things was not what I concerned myself with when I was in architecture school and now I'm kicking myself!
-
@rickgraham said:
Unfortunately, Mac users are out in the cold with this software as they only have PC.
Rick
it is amazing how some software developers do not realize how much they are losing by not porting their apps to mac. we architects are dying to find an affordable alternative to giants like autocad, archicad, vectorworks and the like but the developers of cheaper and less complicated apps either stick to the windows platform (this is for you, DoubleCad people) or, when they opt for mac, they not always produce competent stuff (i am here thinking of PowerCadd which could be an alternative but has lousy and insufficient documentation: not a single video or tutorial is to be found in their site).
I apologize for the ranting, guys.
-
@unknownuser said:
Very good comment you are making. I think what made these "old School" production methods work so well is that you were intimately aware of what you were doing.
Well yes you were but is a drawing (what you were doing) what you really need? I mean you used to need a cheque to draw cash from the bank. Now you can use an ATM. By doing so you minimise human effort and error. And you don't need cash cheques anymore.
@unknownuser said:
With all this technology we have lost awareness of what we are actually doing.
I think the awareness question is more what we want to achieve - the creation of a building or the creation of a beautiful set of drawings?
-
Chris,
In short I was making two observations about the role of technology in work produced and might have muddled them together.
- There is a lot of "tail wagging dog" going on. That is, programs forcing the designer to do what is easy on the program.
- To modify the program to your needs is a whole new and largely unanticipated project on top of existing projects. (BTW people that participate on this forum are largely in contradiction to this statement. Way to go guys!)
-
@unknownuser said:
- There is a lot of "tail wagging dog" going on. That is, programs forcing the designer to do what is easy on the program.
Yes I quite agree, and more than this technical prowess can overshadow real creative achievement (something tangible rather than more general "creativity") - looks good must be good kind of thing.
@unknownuser said:
- To modify the program to your needs is a whole new and largely unanticipated project on top of existing projects.
Also agree and have suggested re-looking at the whole design/build/use/recycle process bearing in mind what technological advances have been made say since John Walker started Autodesk - Internet communication and software, 3D modelling, animation, file sharing and so on. Nearly two years ago I posted this short video which you may find interesting or disturbing.
-
@edson said:
or, when they opt for mac, they not always produce competent stuff (i am here thinking of PowerCadd which could be an alternative but has lousy and insufficient documentation: not a single video or tutorial is to be found in their site).
I apologize for the ranting, guys.
Edson, PowerCADD has its own particular problems. I think its main problem is the primary developer. Otherwise, being a good 2d drawing program, there isn't much to tutor or make a video about, although third parties have made some attempt to fill in the gaps. The WildTools (extension) manual isn't bad. I used to do one-day introductions for PowerCADD and other than a few phone calls or emails, people usually took it from there themselves. PowerCADD just hasn't evolved a lot over the years, nor has their web presence or advertising. Now PowerCADD has a new export system that means you can simply copy and paste into 3D programs like Bonzai3d. It could work for SU too if a programmer wished to look into the apparently simple (xml) plug-in development. (And any other two programs with the plugin can also share)
Namesets is intriguing. I am reminded a bit of the Condoc system that was more of an attempt to augment and organize drawing data than a whole new approach. At present I just don't grasp very well, myself, how Namesets will translate into a design and communication medium, but I hope to see more.
-
@pbacot said:
Namesets is intriguing. I am reminded a bit of the Condoc system that was more of an attempt to augment and organize drawing data than a whole new approach. At present I just don't grasp very well, myself, how Namesets will translate into a design and communication medium, but I hope to see more.
As inventor of Namesets I should perhaps offer to show you more but I'm a bit stuck at the moment for time (we are moving back to Hong Kong involving prefabrication, transport and long distance renovation initiation and supervision) and also stuck for normal Internet resources (the most useful being blocked). However my web site is still functioning so I am told, and I have also been told that "And then there is Operative Content" is my best shot at explaining Nameset design and functionality.
But also I would like to quote from Edson's web site which in my view is fundamental to this topic.
@unknownuser said:
Our studio aims at producing adequate and consistent architecture, related to the essence of its time and place. The notion of adequacy –or pertinence– is fundamental to us. In a period when there is no consensus as to what beauty and architectural quality should be, the only goal we can aspire to is the creation of adequate formal/spatial solutions to needs that require spatialization. Adequate form is that which synthesizes – by way of a formal structure that could be visually understood – the demands of the program, the suggestions of the place and the discipline of construction.
How different from mountains of Awesomeness-ness. Three additional points occur to me in respect of adequacy.
... in terms of compatibility an adequate solution is HTML and Javascript. Well not quite yet; it seems we need Sketchup in its current form until HTML 5 is matured. The sooner we get rid of the idea of proprietary software and treat it as a utility the better.
... adequacy of communications relate to both content and format. A beautiful render required (adequate) to sell the design is totally useless to the setter out who is quite happy with a few clear 2D lines and levels. But both should and can be generated separately from the same set of data.
... adequate competition is not lump sum competitive tendering nor any of its variants, but the result of deep and yes transparent interaction in an up-to-the-minute dynamic marketplace.
-
I have become a little obsessed by the notion of adequacy and want to try and explain it, at least to myself. It seems to me that if all aspects of a project are adequate the project can be said to be excellent; and if one aspect of a project is more than adequate another must be less. Can that be true? Using "pertinent" in its stead, "The spaces are more than adequate but the cost is not pertinent to the budget."
But adequacy is the real test. There is no point judging adequacy with square foot allowances; it has to be tested using collections of models of real objects right from the beginning. For example, an office floor can be first represented by a number of workstation models; each model is tested for adequacy; then each grouping, then the whole floor including circulation and services and of course site restrictions. Very quickly, the assumed adequacy of square foot allowances is queried eventually (hopefully) to become extinct.
This kind of thinking lead me to pursue Namesets and Sketchup as BIM.
-
I have seen references to "Rhino" on the forum in the past, but have never taken a serious look at the system. It seems they are working on modules for BIM-like functionality and have released one module for structural work. I saw their display at the AIA convention in Miami last week but didn't have time to look into it and their display was rather confusing with all the "Zoo" references for their multiple modules- I found it to be a bit overcomplicated and put-offish.
I did get to talk a bit with Rich Hart from Render Plus who was sharing the Rhino booth area.
Anyone have any recent experience with Rhino with the BIM slant? I just wonder if there is something to be learned that might transfere to GSU
-
I think you'd also need to focus in on what the target is - BIM for what? Houses? Hospitals? Airports? Commercial mixed use / retail? They all have their own special languages...
And while I'd agree that many technologies and processes can scale - the language of the residential builder is vastly different from that of the commercial developer / project manager. Essentially one needs to codify a language for the process of communicating what you want built - and its huge in terms of scope when looking at how to create a generic system that could work for multiple projects / types / scales. Add in software guys who most likely have never swung a hammer or climbed scaffold - and you create several translation problems - the first being codifying the language of building so that most folks agree on what the heck you are talking about, the second getting the programmers to understand how this should work in software' the third getting the software to be logical and referential to 'the real world', and fourth - getting the end user (designer, builder, architect, engineer, etc.) to understand how to make the software go - without having to adjust or adapt problems from the real world to fit in the box the software painted you into.
The software cycle is pretty quick - buildings (should) last a long long time. The language of building has been morphed from that of master builders into professional categories and subdivisions and specialties from the architect on down through the mechanical contractors and the day laborer carrying the concrete forms to the hole from the truck.
In the end, its about the building.
Tools continue to change. And with the fast cycle of software - it seems that the tools may always be lacking when trying to virtually control all the aspects of building in a physical world.And in regards to 'adequacy' - I'd rather be effective than efficient than adequate any day.
-
@bmike said:
And in regards to 'adequacy' - I'd rather be effective than efficient than adequate any day.
Just on the last part as it is near time for bed.
That was my first thought ... yes effective and and efficient seem better words ... but how do you test them. Adequate is capable of reasonably simple and objective empirical analysis and could perhaps become as acceptable as "just in time"... which prior to 7-Eleven I remember would have been viewed with derision.
Goodnight.
-
Sorry to jump in so late on this, but I overlooked it thinking perhaps it was another "we wish SketchUp could" topic.
Instead I find an excellent dialogue.
So to actually accomplish what you are discussing, what would be necessary. Is SketchUp really only a collection of faces and edges, which in essence is mathematical representation of a 3D object.
If this object is a wall, how could you enter all the data that is necessary simplistically enough to mirror the SketchUp user friendliness, so that this data could be extracted by anyone using the same software to apply the information to their discipline?
This is a big undertaking, needing to codify everything from its position and location on the planet to the data needed to understand, quantify, and cost out etc, all the properties of this wall.
So would I still be able to paint a texture on this wall, or would there have to be a whole new library of materials that have embedded the information of the make-up of the materials.
Does this mean that every time I do something a dialogue box appears with blanks for me to fill in? If so would this take away from the SketchUp ease of use? If not where is this information coming from?
What would the nuts and bolts actually be?
This is a very big topic, and a very interesting discussion. -
@bmike said:
I think you'd also need to focus in on what the target is - BIM for what? Houses? Hospitals? Airports? Commercial mixed use / retail? They all have their own special languages...
A long time I talked to a professor about Boeing's use of solid 3D engineering to design and manufacture their Boeing 777. Here's an extract from Wikipedia's article:
@unknownuser said:
The 777 was the first commercial aircraft to be designed entirely on computer.[18][24] Each design drawing was created on a three-dimensional CAD software system known as CATIA, sourced from Dassault Systemes and IBM.[31] This allowed a virtual aircraft to be assembled, in simulation, to check for interferences and to verify proper fit of the many thousands of parts, thus reducing costly rework.[32] Boeing developed their own high performance visualization system, FlyThru, later called IVT (Integrated Visualization Tool) to support large scale collaborative engineering design reviews, production illustrations, and other uses of the CAD data outside of engineering. [33] IVT is still active at Boeing in 2010 with over 29,000 users. Boeing was initially not convinced of the program's abilities and built a physical mock-up of the nose section to verify the results. The test was so successful that additional mock-ups were cancelled.[34]
The prof said he had discussed it with his post grads in relation to the building industry but they concluded the parties involved were just too diverse. This idea still persists bolstered by the industry's self interested conservatism. Despite technological advances today's contract documents are really not so different to those I made in the 1970's for chrissake.
What actually allows these interests to meet are the physical elements of the building. The labourer collecting the forms, engineer calculating rebar, buyer purchasing concrete, all in part derive their living from a single object of interest like a column.
So if using the name of one of the projects you mention you can immediately access lists of things you need to consider, then use these to get more until you have a suitable hierarchy to identify and locate purchasable objects you will quite quickly assemble a model of the project made entirely of words - an associative index if you like to access information about what each word represents and a means to assemble a model ... or the real thing. (SU-wise the information can include component file name and transformation)
@bmike said:
... Essentially one needs to codify a language for the process of communicating what you want built - and its huge in terms of scope when looking at how to create a generic system that could work for multiple projects / types / scales.
I think when you make a scheme based on physical objects language difficulties are much reduced. The names of such objects live in hierarchies and have SU models to support them so, for example, there might be a sofa, coffee table and TV collected together under "Living room" or beams, columns and slabs under "Structure". Yes it is huge but this is not about a single software company but an ISP type means to collect experience and redistribute it, a bit like file sharing (where the index is a glorified playlist). It is a scheme that could be started on a single project with demonstrated benefits of consensual advertising encouraging it to evolve.
@bmike said:
Add in software guys ...
All software guys need to do is to provide a simple Internet enabled "machine" (see static example in scrapbook linked below) to build the animated hierarchies of names from lists from industry-biased digital markets, and to set and retrieve plain text records.
I don't know if this is a satisfactory response to your thoughtful post. But whilst the idea is simple I tend to rave on a bit ... so I will stop here for now.
-
@dale said:
So to actually accomplish what you are discussing, what would be necessary.
Some professional to make the nameset machine from my performance specification/amateur coding.
A design firm willing to help run a trial project in parallel with their normal routine.
Models of products to be used in the trial project.
Funding for the machine, trial and models.
@dale said:
If this object is a wall, how could you enter all the data that is necessary simplistically enough to mirror the SketchUp user friendliness, so that this data could be extracted by anyone using the same software to apply the information to their discipline?
One of the principle functions of the machine and its devices is to reuse names, so that input is minimised. Even when used unique values are added as new prompts/options.
I find the prototype machine does not interfere with the Sketchup interface; in fact because of its small size and its ability to turn components on and off in different combinations, it makes scene making much less complicated and uncluttered than messing about with the layer inspector and outliner. It also uses observers so even now you can manipulate models directly or with the machine.
The machine and devices are made with html and javascript. Imported prompt/options and project records are stored in plain text. I believe this is about as interoperable as you can currently get.
@dale said:
This is a big undertaking, needing to codify everything from its position and location on the planet to the data needed to understand, quantify, and cost out etc, all the properties of this wall.
Every product whether made in a factory or on site would come with its own machine manipulable datasheet. For example, say there is a one brick wall component that is 1m x 1m. Its plain text datasheet details brick reference, dims, cost, delivery terms, mortar allowances and so forth. These would be part of the minimum registration requirement.
This type of component would be automatically made unique and sized and positioned manually or with machine devices that also enable setting design criteria.@dale said:
So would I still be able to paint a texture on this wall, or would there have to be a whole new library of materials that have embedded the information of the make-up of the materials.
I don't think presentation type things like textures need be affected, but in my view it would be best if manufacturers realistically textured their own models. This of course raises the question of modelling disciplines, file sizes and so on, which I think can wait.
@dale said:
This is a very big topic ...
It is but it can start small as I wrote above. Machines will be freely available, with storage of records determined by the machine owner. I haven't thought too deeply about the registries but I am sure there are better people than me to look after that. The main point is that this is a scheme that widely distributes the burden of compiling small amounts of data.
-
It appear to me that many are wishing they could assign properties to SU objects so they later use that for documentation. Is there really no such tool out there already?
Advertisement