Keynote session A: ONTOARCHI – an ontological framework for developing a semantically based new generation of automation tools for architectural design
This is a modal window.
The media could not be loaded, either because the server or network failed or because the format is not supported.
Formal Metadata
Title |
| |
Title of Series | ||
Number of Parts | 13 | |
Author | ||
Contributors | ||
License | CC Attribution 3.0 Unported: You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor. | |
Identifiers | 10.5446/48219 (DOI) | |
Publisher | ||
Release Date | ||
Language |
Content Metadata
Subject Area | ||
Genre | ||
Abstract |
|
5
6
7
8
9
10
11
12
00:00
ArchitectComputer animationLecture/Conference
00:36
Spirit levelBauträgerLecture/Conference
01:14
ArchitectLecture/Conference
01:54
ArchitectureLecture/Conference
02:37
Cut (earthmoving)ArchitectureLecture/Conference
03:13
Typology (theology)Lecture/ConferenceMeeting/Interview
03:57
ConstructionEarthworks (engineering)ArchitectLecture/ConferenceMeeting/Interview
04:44
ConstructionArchitectSpaceBauträgerArchitectureLecture/Conference
07:04
Scale modelArtificial turfMeeting/InterviewLecture/Conference
08:04
Artificial turfMeeting/InterviewLecture/Conference
09:14
CentringProfilblechLecture/Conference
10:10
BauträgerProfilblechTypology (theology)ArchitectureMeeting/InterviewLecture/Conference
12:34
BauträgerArchitectureLecture/Conference
13:14
Lecture/Conference
13:51
Water wellLecture/Conference
14:34
Earthworks (engineering)ArchitectLecture/Conference
15:12
ArchitectureWater wellTypology (theology)Foundation (engineering)BauträgerSpaceOil platformLecture/Conference
20:04
SpaceLecture/Conference
20:46
Water wellArchitectureLecture/Conference
21:26
Air conditioningWater wellArchitectureEarthworks (engineering)SpaceLecture/Conference
24:05
Issue (legal)Lecture/Conference
24:51
Lecture/Conference
25:31
SpaceProfilblechLecture/Conference
26:17
ArchitectureLecture/Conference
26:54
Interior designSpaceArchitectureWallVolumeLandscape architectureSpirit levelRoomArchitectConcreteDoorWindowLecture/Conference
32:03
Lecture/Conference
32:42
Earthworks (engineering)Global cityArchitectLecture/Conference
33:18
Civil engineeringArtificial turfArchitectureEarthworks (engineering)Water wellSpaceLecture/Conference
36:18
Meeting/InterviewLecture/Conference
37:25
ArchitectMeeting/Interview
38:12
Lecture/ConferenceMeeting/Interview
39:14
Portal (architecture)ArchitectureEarthworks (engineering)ArchitectPile driverSpaceMoldBuildingMeeting/InterviewLecture/Conference
45:11
Issue (legal)SpaceScale modelMeeting/InterviewLecture/Conference
45:50
SpaceIssue (legal)AbwrackungLand lotLecture/ConferenceMeeting/Interview
48:03
Computer animation
Transcript: English(auto-generated)
00:00
This is a presentation from the research team that organized these symposiums. We are working since the late 90s, since 2011 began those symposiums, and something
00:35
we are trying to do is to produce software with commercial quality to the architects
00:48
in some areas, some fields. Those tools are lacking because they are developed in the research.
01:10
They are in the development level that are not very prompt, very keen to be presented
01:24
to a large public of architects. That is our intention. We produce some pieces of the software. The last one is this one, is something that there will be some discussion about this software.
01:52
I am not developing this thing. What I am trying to do is to present a new research field we are trying to develop in
02:07
our research team. It is something that we call ontology based design, design automation to fulfill the tasks
02:26
of architectural main task, creative task, and not those that are present in current
02:42
software aids to architects. The main problems we can encounter in, you can see in actual computer aid design software,
03:02
cut software, commercial off the shelf, I say the current types, the current tools in computer aid design is that they are not directed to the creative part of architecture.
03:20
The first, I would say there are two generations of computer aid design tools. The first one is something like this, and this type of CAD were directed to help drawing.
03:44
The second generation is, well, you can see there an example of those type of CAD. The second generation is directed under the revolutions, for example, BIM.
04:03
They are directed not to the creative part of architecture, but to the construction tier of the work involved in the construction industry.
04:25
Architect has a few instruments to develop with formal instruments and digital tools
04:42
to develop what is, in fact, the profession of the architect, and it's not drawing and it is not construction, but is the fulfillment of human and social needs in the construction
05:10
and the environment. The architect is the guarantee of those uses of the construction that fulfill the needs
05:36
of the people and societies.
05:41
One of the greatest of those problems is the organization of the human activities. Well, none of those problems is addressed by those two generations of CAD tools.
06:05
Well, those are examples of developments of CAD tools for BIM, BIM and the second generation CAD are indeed very helpful instruments, but not to the creative part of architecture.
06:34
So there are very many narunas tools, some of them are expressed there.
06:46
There is no knowledge involved in relation between social and human function of the spaces and the definition of those spaces.
07:05
Those relationships, what I could say, architectural theory are not expressed in those computer design tools. Well, I'm not going to develop those topics I have there.
07:35
Well, now I must arrive to what are our solutions.
07:47
Well, what we think will be a solution is what we can import from intelligence, artificial intelligence.
08:00
This is a very confusing model of what is an agent for artificial intelligence. I have here a much compact scheme.
08:23
An agent for artificial intelligence must have a process of knowledge that is obtained either from the world, from the knowledge of the world, either from the introduction
08:43
of something that is already known and people can bring to the experts, can bring to the artificial intelligence.
09:00
When we have values, this is very important because there is no way to raise decision only on knowledge. We have to introduce values, human and social values that determine what is the direction
09:21
of the action. We have to introduce knowledge and values to take decision process and that is where it's in the center of that scheme there.
09:41
And then we must take actions for the architect, the action is in the form of projects in the process of design. Well, I'm passing there.
10:01
We will have time to discuss more in a few moments what are our guess that will be the development of a third generation computer aid design tool.
10:22
Our guess is not very far from the development of Dreamcatcher, Dreamweaver from Autodesk. But there are a few differences. The first is the definition of knowledge in an ontological form.
10:50
Ontologies are descriptive language, they don't act but they can increase knowledge
11:04
capacity in a descriptive form. We can describe our knowledge on architecture in the form of ontologies. Ontologies are very formal language based most of them in the near first logical, first
11:23
order logics. They have some advantages over other types of logics because they can accommodate, they can accommodate contradictions although the most known logic don't accommodate for contradictions
11:55
and the development of ontologies they can accommodate. And that is very important because ontologies are the process of many actors interact in
12:14
the development of knowledge. You can say that in what we're trying to do, there are many interventions with very
12:28
different goals and they have to collaborate but their collaboration does not need to be completely perfect.
12:41
There can be some disturbance in that collaboration. Ontologies are very akin to that kind of development. So it is our guess that this is the way that we can develop the depository of architectural
13:14
theoretical knowledge. There are tools that are well current in ontology domain and we decided to go for them.
13:31
The language is where ontology language known as O. There is a development environment from the University of Stanford.
13:43
And there are many other tools that we can, well they are current in the community that deals with ontologies that can provide a certain degree of well known, they are already known
14:06
by many people, there is a critical mass that can develop with this technique. Well second is it's not enough to have a descriptive language.
14:30
We must act with that language. Well the second tier of our research is developed not only a language but reasoning or reasoning
14:48
tools that can create departing from that knowledge that can create what we want. That is the project of the architect, not in its final form, of course, but in well
15:11
helping the work of the architect. Those reasoning tools are also very formalized in ontologies.
15:24
So we can develop those reasoning tools with well some soundness because it is formalized and a great openness because they are very powerful.
15:49
So it will be, well it's the other aspect we are trying to develop. Well there are some kind of potentialities that we could see in the use of ontologies
16:12
and there are some other fields that can be developed not only in the project or design
16:23
process but in all those aspects that you can see there in the slide. Well our goals in this project is, as you can see there, a set of products, foundations
16:44
ontologies for architecture and urbanism, some subdomains, design automation, integration with well-known tools, BIM and ISCs, interfaces with well developed tools for engineering
17:11
and environmental structure and so on. Interfaces with well developed scientific field in architecture, space syntax and many
17:26
others products. You can see there that the goal is not only the automation tools for the architectural design although that is the main goal of the project but also there are other fields
17:49
where ontologies and reasoning tools over those ontologies can be developed also.
18:00
Also an institutional platform and this is why we are presenting here this project because we need partners. This is a gigantic task as you can see and only with collaboration between many research
18:30
teams that can be taken to a good end. Of course this is not the first time the problem is addressed.
18:45
We are not inventing these ontologies but we intend to collaborate with many other teams in these objectives.
19:09
This is a design of ontology. We developed already an ontology for architecture. This may be very confusing because there are abstract concepts that are correlated in the
19:28
ontology. I have here the main concepts that we are trying to deal with and those are, I'm sorry,
19:50
this is in Portuguese, I must translate to English because I think this is important. The first place is the type of persons, of people.
20:06
They have activities that have processes that are sets of movements. They practice some schedule, they have some environment features and necessities and
20:30
they need equipments. All these have some requirements, environment requirements. I would say the most important in our analysis is space but there are many others like accessibility,
20:49
visibility, noise, lightning, security, some characterization of air, quality of air.
21:03
There are people that live in an environment. That environment also has some special characteristics and some environmental characteristics. Constructive elements have also some environment characteristics.
21:23
Well those are the main concepts that we developed for the ontology of architecture but it's you can guess there will be, there are already developed hundreds of other concepts that
21:46
are related with this. What these try to say is, well, these global scheme is that people act in an environment.
22:03
They need space. They make movements and they need some environment conditions. And the space and the elements of, and constructive elements can provide that space, that environmental
22:27
conditions. Well, when we say that we need something for people to use, we can obtain if we have that knowledge in an ontology, we can query ontology and obtain the desired space and the environmental
22:57
characteristics or features that must fulfill the desires of the work of people living in
23:06
that environment. Well I want to show a small example. We are still in our very beginning. It is a kindergarten.
23:23
Well I must say that this project is not at the very beginning. We are working already with some other schools. We indeed tried to get some funding from Horizon 2020.
23:45
We made a project. And there were some different approaches to the creation of ontologies. Our solution is what we call a neurotic solution.
24:02
We go to the already known architecture and here we are only dealing with the space needs and there are thousands of books and papers on this issue.
24:29
So there is not yet formalized but there is a very big research on this team.
24:42
Well I think everybody knows Neufert. You go to Neufert to know the dimension of the kitchen or something like that. Well the research is centennial and there is a very big effort in this kind of problems.
25:14
There are other approaches for ontologies. As you can see our approach, we are dealing with this kind of approach because there are
25:27
others. This kind of approach is heuristic but there is another kind of approach, heuristic that they are associated with the name of shape grammars. In the case of shape grammars, it's a little different from this because in the definition
25:45
of the ontologies, they have already actions to create design. Our approach is there is no action, only declarative what we put in the ontology.
26:06
In the reasoning tool, yes, in the reasoning tool there is actions to produce the design. But space grammars already introduced in the ontology through the form of rules of production
26:23
of the grammar of shapes and it's a little different. For example, the University of London is working on that solution.
26:40
But there are other solutions, solutions that are non-heuristic. In those solutions, we don't go to see what already is known in architectural literature but we go to examples in architecture and we try to understand, well, understand this
27:02
and discuss because there are processes, ontologies that we retire rules without perfectly understanding what is under this knowledge.
27:27
They are very used in everything. and non-heuristic trends to arrive to ontologies or to knowledge without knowing very well
27:51
what understanding, the knowledge is not based on understanding but in other correlation of variables.
28:02
Well, our solution is we go to the biography, to the literature of architecture. We retire what is known, for example, what is needed in the garden. The spaces that you need, the dimensions that you need, and then you put that on the ontology
28:29
and then with reasoning tools, very, very abstract and general reasoning tools, we can
28:46
make a distribution of spaces and the interface between those spaces. Well, that's one of the levels of the design of the kindergarten is the whole location
29:03
with interior and exterior places and the relations between those spaces. This is the interior of the kindergarten, the spaces with, well, shapes are not yet
29:23
developed in our model, only volumes and you can see in the scheme and the relations between volumes. Well, the main variable to optimize that we are dealing with is accessibility, but many
29:44
others will be introduced to define the reasons why use some shape, why this room must be near that room, why there is an interface between two rooms, that is a wall or something
30:06
else or a door or a window or what characteristics of the constructive element must be produced to fulfill the environment problem that we are dealing.
30:23
Well, of course, we are dealing with real examples and what we are trying is to create a reasoning tool that can be generalized to create any kindergarten. We would say we need a kindergarten for 15 or 100 children, how would it be like?
30:52
It's not very different what you see here from what you see here. This is a scheme that represents more or less what is a real kindergarten.
31:04
It's not by chance because we introduced the data that created that project in particular. Not a project, but many projects. So when you can it's not surprising that we obtain the results that are the common practice
31:29
of architects. What we are trying to do is not a better architect, but it's formalizing what the architect does.
31:46
Nothing more than that. Well, this is the third level, the interior of the rooms. The distribution here is much more particular. We must make the distribution of, well, concrete pieces of equipment needed and, well, here
32:13
it's very much more difficult way to arrive to a good solution.
32:28
Well, I'm ending here only to say that what we are trying to we are not trying to end
32:41
the scheme, but to transform then the scheme to real project. And there are already tools. This is another part of the global set of works that we must do. And well, this is well, several works.
33:03
One of them is surprising from the 60s from a well-known architect in Portugal when he was very young in the 60s or last century is Nuno Portas with some well, they were the
33:25
fathers of artificial intelligence in Portugal. They worked in national laboratory of civil engineering in the time. And they produced well, from schemes to real design of the project, of architecture project.
33:51
This is not from them, but it's from the third symposium we had well, two years ago presented
34:01
by the subscribers of the paper, something that I can't spell. Well there is also optimizing distribution of spaces to create the optimized well, beginning
34:28
in the sketch and trying to create a real project, architectural project.
34:41
This is not the this kind of work is already well developed. What is not very well developed is the knowledge to create the scheme. And that is what you are trying to develop.
35:00
Of course, the final goal is to arrive to the capacity of exporting all these to complete projects. Of course, always with the intervention work of the real architect, because it's impossible
35:24
to create projects without the pure automation, without the intervention of the real architect. And the final result of IOGO is the creation of BIM, B-I-M, projects with the help of what
35:48
we are trying to do. Well I'm sorry, I'm open to your questions. And more than that, well I'm hoping that some of you would be interested in participating
36:11
in this effort and go to further collaboration. Thank you.
36:46
No question I would say that there will be a question, because this is so controversial
37:14
problem, the automation of the architectural design, that there must be a question there.
37:26
What do you expect will be the reaction of architects to this tool? I understand that this could be very useful, mainly in large projects and complex projects, for instance hospital, which is so complex that this could really help.
37:45
But what do you think is the reaction from architects, which are beings that know everything, and what do you think they will react to this interaction between this software, this ontology,
38:03
and the creative work of architects? What will be the reaction of them? Do you want to respond now or any other questions?
38:20
Yes. Victor, I can see that on your diagrams, especially with the circles and the rectangles, it's very clear that programmatically it's completely solved. I would like to understand if there is some way to analyze the spatial qualities of this
38:40
kind of organization. Well, I think that a full automation of project is beyond our imagination. We cannot think of that.
39:00
But I will follow the engineering process. Well, I am an engineer, I must say. I have a very known engineering portal already that is of engineering that made some of the
39:28
good examples of engineering in portal. When I work with him and I try to implement computational studies of his work.
39:48
One of the greatest buildings here in portal, it's the tallest building when he constructed that building, he said, I spent six months calculating this building.
40:05
And now I was on the computer, you make the same thing in six minutes. When I finished the project, I said, what an engineer I am. And now in six minutes the machine makes the same thing.
40:21
Well, of course the engineer is not dispensable by the machine. What he makes is another thing, not sums and multiplications, but the creative part of
40:44
engineering. He does not have to make the calculations because the machine is much better than a person to make that. And what he makes is global design of the structure.
41:01
And he thinks on what is really creative. What I would say in architecture is, well, I suspect it will be the same thing. The architect does not need to be aware of everything in architecture because there are
41:20
some things, many things that can be made without the intervention. Because it's something that is well known, the process of creating that design is already known. There is nothing new. But there are many things that are new and that is what the architect may produce, exactly
41:49
the creative part. What is not creative, the machine can do well. I think this is a normal path of everything.
42:04
Well, now an automobile has many digital, the driver is aided by many digital tools in the car. Well, if you want to be creative, it can be, but well, those tools are very, very important.
42:29
And I think in everything, this is the path we are making in every field of our activity.
42:47
It's what I think. Of course, in our solution, there will be many things and there are parameters that in every stage of our process, what you're trying to do, where the architect will introduce
43:12
what he feels, what is the feeling, because there are many variables that he can control
43:21
in many ways. And there are things that there will be never a solution, a formalized solution, is what I think. Relative to your questions, well, we are trying to, for the moment we are dealing with volumes,
43:46
quantity of space, but not with the shapes for the moment. We are not dealing with the shapes, but although shapes and the variable that we are dealing is accessibility, but there will be many others in the beginning because, well, I would say,
44:07
and I defend this, the main problem in architecture is what the architect makes is conforming the human life in space.
44:23
It creates spaces for people to live. And what the architect do is molding space. I think you understand me. Architect mold space, and the space is molding life of people.
44:44
Of course, the universe may be more true, even life of people mold space. It's a correlation between the both. Well, we are trying to deal with human activities that are well established, and you can read
45:09
in many, many books. There is no controversial issues in those things. And it's what we are trying to do now.
45:20
But shapes, yes, they will be another issue that we'll deal with later. Well, for example, space syntax can deal with some characteristics of quality of space.
45:44
And we are trying to implement linkage, some link from space syntax with this model. So quality of space will be an issue that we are trying to deal, but not for this moment.
46:08
I would like to highlight this relation between these two questions. I think they are really important for the rest of the symposium. It's how to establish the relation between creativity, quality of space, and what we
46:23
think about quality of space. What is quality of space? This is not a linear discussion. This is not something that we can establish in a very clear equation, a mathematical equation, because quality of space has a lot of inputs that are not clear to, or not the same to
46:43
all of us. So this is one of the main goals of the symposium, as you will see during the days. And that's why this symposium is not about a specific approach, a specific method, but we are trying to understand how we can correlate all of them to achieve this, achieve creativity,
47:06
design, achieve quality of space, and to get the best of each methodology despite the limitation that all of them have. So before we go to the coffee break that is already set, if no one has another question,
47:25
I would like to stress this triangulation, this correlation between these three main issues that will be all together during these two, three days that we be together here.
47:40
It's creativity, quality of space, and formalization. How do we formalize this? How do we formalize quality of space, and how do we formalize creativity? So this is one of the things that I think it will be as common ground for our future discussions, and I hope you get it in mind.