Player is loading...
The Desert Rose - When poesy meets technology
Nassim Saoud, Trimble Consulting, Paris, France
Tuesday, Oct. 29, 2019 · 4:30 p.m. · 22m 48s
Envelope design optimization - Project federated database and analysis - Constructible assemblies
Embed
Transcriptions
Note: this content has been automatically generated.
00:00:01
hi everybody men in business him so damn trains an architect uh working
00:00:05
for good technology that have been remanded remote control things um one of the
00:00:09
a basic care some one of the directors or seen our operation um in europe and um at least
00:00:17
just a quick background about where we came from uh so get a technology a it was um i'm a
00:00:22
born in a design practitioner lay in a friend gary office we've been playing the role of the r. and d.
00:00:28
groups and early to tiles and uh we spin off
00:00:31
after serving a exclusively only frank a project was not working
00:00:35
with frank eco systems every every project engage engine use
00:00:39
builder's subs and so one was not working also with other
00:00:42
uh designers and uh uh we always have a kind of from very tied i'm willing to that
00:00:49
tack industry so that's soul was one of the uh shoulders in good
00:00:54
technology back in early two thousand although that's short for a period of time
00:00:58
and says two thousand fourteen we've been acquired by trembled so remote consulting is getting technologies
00:01:08
so uh what we learn out of a frank track is if i
00:01:12
could just summarise it is just like the management of complexity that was substantially
00:01:17
geometry so uh we've been uh also looking into other industries a substantially them
00:01:24
transportation aerospace so we borrow talons after we we borrow technologies
00:01:29
the the same that are used to uh to make aircraft
00:01:31
and we recycle in them in order to adapt them to
00:01:34
the specificity of uh the construction industry so the learning of
00:01:39
of the building side uh was uh we've been able just to to deploy this is the kind of snapshot what
00:01:44
have been uh all the project over the past two d. k. that we're gonna plan them into different type apologies
00:01:50
a high rise building a kinda make a infrastructures including
00:01:54
the linear infrastructure that have another set of of complexity
00:01:59
this is like some of the uh cutter buildings we've been involved in
00:02:02
uh some of them design by frank some of them by novell or or by others
00:02:07
uh uh our paths cross with the m. add some multiple times uh you've been involved in some of those projects
00:02:14
uh it's i'm sure it is because the timeline was a
00:02:17
very interesting so we've been tackling sometimes multiple project on the
00:02:20
same period so uh in a in somehow for that's when we we don't and got that much you're using have been raise
00:02:27
like a uh like to doesn't seven thousand eight uh altogether is so we've the those present a really specific
00:02:33
the cutter building i i personally consider them as a laboratory with open sky so each of them was really
00:02:39
the excuse just to explore and just put the limits of the enjoying a a filled
00:02:44
uh we don't for example what the maternity of thirty one peyton twitches pew
00:02:48
innovation for the construction industry so some of this project even exceed the uh the
00:02:52
only requirement and uh it was really great uh to be involve on them so
00:02:56
today we gonna talk more specifically about cut time machine museum one on the middle
00:03:01
and i did is a very uh simple just side by side to to
00:03:04
uh to put them all together so generally um uh cultural uh building on these
00:03:10
or a scale that like in a similar size but i was in somehow uh luckily
00:03:15
bit much more bigger uh uh in terms of scale and even in terms of a complexity
00:03:23
uh i'm not a talk i mean uh and to which um about the design specifically
00:03:29
but um i do that i this lecture was interesting for me because i i have to
00:03:33
look backward so we spend plus seven years on this project so i just to look into
00:03:37
many people try just to put things together and just give you today uh kind of outlook
00:03:42
i'm sure we do mainly the back of house you know the cuisine be half of a a project um
00:03:48
what would this complexity so the initial intent was uh wintertime middle east
00:03:53
so there was a like a beautiful metaphor to go through and um look into mineral does does the roles
00:03:58
but the initial as scheme that was presented a by uh john well
00:04:03
was a like a like a basement buildings um uh below grade that was
00:04:07
you know just to protect the visitors and the all the piece of art inside
00:04:11
so uh for those of you know the middle uses a very aggressive climates or temperatures could just exceed
00:04:17
fifty five uh the reason that the first keen but then the client um
00:04:21
had of um can make at i'm using the target you and the uh princess
00:04:26
uh this is sort of the crown prince over there so was look into
00:04:29
something uh much more i conic just to showcase you know them qatar uh
00:04:36
past qatar present in qatar future so that's what the museum is about just explain
00:04:41
um the history of um um uh of of
00:04:45
of guitar as a nation and then all the major milestones uh with the uh
00:04:50
um discovery of owning as excel trying to kind of forward looking so it's like
00:04:54
that's a continues corridor with different accent that's how the building um is made
00:04:59
the inspiration what the there's the roles at the start by donna well but the engine was not just
00:05:03
to them may make exactly uh like this so the initial scheme was mainly playing with our example disks
00:05:10
and then when we started to get to construct abilities instructors we've
00:05:13
been seeing more more ablaze in vertical disk and that's how we get
00:05:16
uh back to these like uh uh shapes of uh uh there's a rolls
00:05:22
so i'm i'm gonna just try to to go fast and used to to
00:05:26
uh to save time because uh our and it's uh the the last i'm sure
00:05:30
you guys i've been seeing a lot of good stuff over the last uh these days so the program is quite true h. there
00:05:35
is exemptions base public space teachers uh uh uh uh restaurants cafes
00:05:40
and so on so the very interesting a set of program inside
00:05:44
so this is more or less a time line so we get the privilege to
00:05:48
be invited by a genital region where right at the star one this project was
00:05:51
almost a fantasy so how to do like a organise scales altogether so when working
00:05:55
in a a a day early design stage uh from paris with the a. g. n.
00:06:00
and we start just looking uh to work on the forward uh like
00:06:04
that's what the endpoint uh should be we start just with this scheme about
00:06:08
what the technology what the model should serve as a main use case so the first one just to
00:06:13
to try to assess what's the risk inherent to this uh
00:06:16
uh construction a look in we're working extensively on design optimisation
00:06:21
collaboration coordination problem solving a and so on and then we set the list of master target doing
00:06:27
construction and also uh we remove uh uh we had conversation about boss construction in the early stage
00:06:38
so this is more or less a scheme of what our scope for that i'm
00:06:41
sharing with you so this couple for what just divide it in like work packages
00:06:45
one of them was the bin and the subtitle is how to how to for a check collaborative
00:06:52
uh eco system to allow the multi stakeholders spread into
00:06:56
different geography over the word and the to help them just
00:06:59
work together on side by side once about data data was
00:07:02
i think the the core principle how this project was driven
00:07:06
and it extractions so how do you get many pilot data in order to feed like different uh types of consumer
00:07:12
design optimisation design correlation and constructed which in execution so just like
00:07:16
i think you draw those different um uh uh what packages would not
00:07:20
transition so you we give a glance of uh what we've been doing
00:07:24
the initial team was in bed that inside the the the office in paris
00:07:29
the of the um the schematic design with about in paris and later on the d. d. and we prepare the tender package in geneva
00:07:35
later on other call the smallest from the middle east and the
00:07:39
uh one the other product was awarded to a. g. c. and we
00:07:42
is we adapted um the design principles and just to spread them
00:07:47
into these uh uh like continuum chains of value uh with no distinction
00:07:51
uh the method you see the saps the on the wrap the project managers and then
00:07:56
i would finish the the scope would like a limited time to people that kept the history
00:08:01
uh of the project and uh just to ensure they hand
00:08:04
over and the and the transition to the facilities management and operation
00:08:10
so this it was one of the fundamental sketch that we drove right at the start i'm
00:08:14
gonna take it to the details but basically it's like a black and look at data driven process
00:08:19
uh this this concept of federated master model but then if
00:08:23
we step back once that there is the design model and
00:08:26
then there's the construction model and there's a model uh by the owner so we had just to follow up to design influence
00:08:32
a process is to go from one to the other one and that's how we figured everything
00:08:37
um we furniture culture press on this is what i was explaining the relationship between project i
00:08:43
think it's not something unique uh all of your packed is is you guys are running multi project
00:08:47
so this um does solution was developed a metaphor for that's all we've it all and then we
00:08:52
we we injected no further development into it and it was uh uh adapted for a couple machine museum
00:09:01
so uh uh the first thing was like a deep analysis try
00:09:04
to understand what the project is about and try to retrieve some uh
00:09:08
regular semantics so uh the project was just divide it into areas and for uh doing the design
00:09:14
station later on to to find out like a like a like a much more um a smaller pieces
00:09:20
so thus therefore areas have been divided into nineteen zones and so what so this is the first let's say project
00:09:25
or is in which not only look into geology we've been uh also what chain scope of work was to win what
00:09:31
how this part would be contracted and just trying to to define like a data organisation that is a matter to the
00:09:37
project organisation uh this is a quite although a a screenshot
00:09:42
so to be a the databases like plus one tera bytes
00:09:45
so uh uh so we've been oh in a continuous
00:09:49
a loop looking both to windows folder files a extra so
00:09:54
it's like a kitty a bayes model development so a project
00:09:56
is tiles and thousands of files so that require very tight
00:10:00
uh i'm i'm sorry protocol uh to to manage all those um uh files and folders
00:10:08
so this is more or less the sketches how we can back uh how we come with
00:10:12
like uh like um the the policy is that gonna govern uh all the development of the project
00:10:17
so what you see here could give you the recipe off balls like thousands of files each of
00:10:21
them so so we have the zoning it just like a combination so every every uh um components out
00:10:28
of this assembly is is is driven by those convention name i always make this basic can present all
00:10:34
of us sometimes we have missing pfizer not desktop that we need just to put on the right folders
00:10:39
but imagine assigned job or we have a small piece where everything looks together
00:10:43
that you need to fit in a particular for side so the fact that the
00:10:46
commission name is is not only like a protocol but it's a way more that
00:10:49
have have enough intelligence to to tell about uh uh the component and its context
00:10:56
and this is one of the glossary so into pages will be
00:10:59
capable just to describe the the recipe how this project will be developed
00:11:05
um and then one we have this database all built in what is interesting is
00:11:09
that the um the every consumer of this database could put different sets of glasses
00:11:14
and see different types of things so that the client was really you're just on the stand uh in
00:11:19
a compatible way the progress of this project so we've been all watching the same models for many years
00:11:25
and try just to put layers try to understand uh what we're making progress where
00:11:29
we had we had cable so we've been just tracking all the submission exchange between the
00:11:33
project stakeholders but then we have some analogies to to to understand okay we have a
00:11:37
full ready package that have all traces can go to the next wave and so on
00:11:41
so we we just cable also to adapt a our strategy is to
00:11:45
the reality of a a process by the uh the the different stakeholders
00:11:50
so like a k. o. open eyes k. off bits 'cause probably the most inefficient shape so it's still
00:11:57
a building that have a program that have square meters that need to be fit in so would be
00:12:00
interested to the void generated by those this can track in um almost to the decimal uh what's is
00:12:06
the the layout and just to lay billing all that setting just reach in the the the geometry inside
00:12:13
this is a program track and and then we have like different teams uh some people working on
00:12:17
the ms your graph e. on the inside the inner some people in charge of the envelop doubters
00:12:21
and some people in in charge of the goal of the multi layer sandwich
00:12:25
it between the all of them so we had just also just to make fluent
00:12:29
link between those different teams that are using different a piece of starters
00:12:35
so i'm not the only one that have that's it comparing models and he's
00:12:39
been thanks to iceberg so the visible the attractive the eye candy is the geometry
00:12:43
but the intelligence are in bed that are not visible so we start just reaching 'em in information
00:12:49
um does database so every single
00:12:53
uh services uh for from inside horizontal
00:12:56
vertical uh have been just label in which with information so we could
00:13:00
than just this but i'm a display different set of fanatics and retrieve a
00:13:04
highly precise um a puke you won't be over a or b. y. m.
00:13:11
so um and then i this is what i i personally call that frankenstein is simply
00:13:17
so there's a lot of let's say literature is about the bin that isn't a
00:13:20
progressive said going from a low the trend of foreign lexicon but i do believe
00:13:24
that um every user have a specific need so if you need to talk about
00:13:29
the user ready maybe you need only like a light surface model to describe the disk
00:13:34
so it's uh what i call the maltese keller and the hybrid level of detail so
00:13:38
it's a kind of baskets there you could retrieve a low details for still and then every
00:13:43
based on the bow based on the needs everybody could just
00:13:45
customise the assembly that they need to perform their uh their work
00:13:51
so this is how the story start with this sketch over there so where there
00:13:55
was like a like a agree i haven't impressive for a design workshop and uh
00:13:59
i read it might be and satisfaction of designer that always believe it you could make it better so every duty to
00:14:05
to change the angle based on the views and so one is like a like a like a basic uh design cycle
00:14:10
and then when we we had to pick up is the design modern just to apply some mathematical rules so
00:14:15
there is plus five hundred discs and uh we just to ensure that every discuss an extract from this feature
00:14:21
just trying to look to the geometry and and so one so so we just we
00:14:25
size and we shape the um the um the initial models out of a. g. n.
00:14:32
uh it it's iterative it's it looks like a one shot
00:14:35
so there was like a iterative process uh looking either told macro
00:14:38
vision or make provision so this is how we stop allies um
00:14:41
the the layout the treaty layout of uh of of these buildings
00:14:46
um and then we start like another set of of optimisation
00:14:49
so look into what is a unique panel uh just making family
00:14:53
is optimism but by the moving everything's looks random but they
00:14:56
are driven by probably candles that goal uh that visa holders variation
00:15:00
so we've been cable just to to look at from aesthetic uh point of view from
00:15:05
and you can go performance point of you know it from different spectre in other just
00:15:08
to um a lot of the um the designers uh to take decisions uh over it
00:15:14
and and and then i know we've been just having not gonna continue slow to look you know to
00:15:18
i'm a bit of materials and how we cannot uh install
00:15:22
them and what's the fixation for each of them et cetera
00:15:26
um so i i i'm the patterns was a design in um in a graphical way e. g. d. so
00:15:32
that different schemes and then we had to adapt those schemes um the project i'm into a construct of spheres
00:15:38
so uh and then either we go from the panel to the global assembly or we go from the assembly to the global panel so
00:15:44
some of them have been identify and have been display automatically into the
00:15:49
model to find like the most optimum uh uh shapes uh on it
00:15:55
and then uh as you've seen there is uh the disc itself but most
00:15:59
of them are intersect in you know kind of random way so it generates
00:16:03
a way to many typical cases so and then we need just to look to each of them particularly and that we've been just
00:16:09
plain with the real drawn fake jones um uh uh uh except
00:16:13
right just to to manage a certain scale of uh of panels
00:16:18
so this is the intuitive process so we get like the the master surfaces a design by a. g. n. and then
00:16:24
we introduce another fabrication process is another to adapt them adjust
00:16:27
to to realise every single component that are ready so our
00:16:31
so that our client was mainly the machines not like people so for for uh what we call the five to factory process
00:16:39
once again the the project management so we have like a big uh it's a kind of i conic a project in qatar so
00:16:44
there was a lot of attention to see what's going on this region of the words very interesting there is a kind of a
00:16:50
competition of landmark a museum so there was a lot of apathy to see
00:16:54
real time what's going on what's the status of the progress um next address so
00:16:59
a documentation this project it's for certainty that conditional really uses you don't bill out
00:17:05
of today but they're still contractual obligation so it was a real challenge how to produce
00:17:10
handed if not thousands of of documentation that was a like a mass for every uh plastic holders
00:17:16
so we've been channels as well because we build is like a highly sophisticated they
00:17:21
three d. database but then we've been also struck lean how to get to this documentation
00:17:25
so this is bring it back this like um uh the products that ten years ago in the history of technologies
00:17:30
almost like stone age sometimes it's not the same after that we have today terms of computational powers and so on
00:17:36
so we had to do also the challenge the database and we we we introduce what we call the cost model
00:17:41
so if you need only the general scheme you don't need to compute
00:17:44
every single panel otherwise take you hours just to get a section that
00:17:48
so you know those different strategy to get this documentation at the at the end of the game for tendering and so one
00:17:54
you don't uh it when you get this result you get it in cat you
00:17:56
don't even see if it was coming from whatever catchy or or or something else
00:18:01
so and then different set of documentation going to the fabrication layout
00:18:05
for um unfolding the the basics how to run fast to get
00:18:10
to the and uh you know the interior uh every um single
00:18:14
surface have been labelled a mm a ga generational planted original plan so
00:18:21
so at everything was like three d. driven from the database below q. for blasting exeter
00:18:27
so that's what how the beast looks like ones like um like um when you explore the the the sandwich
00:18:33
a multi layers um there's about ten named stoppers plus all the custom tools that have been build that need to be built
00:18:39
together uh from difference k. would look at it into like
00:18:43
zooming toes on or from the uh the the project perspective so
00:18:48
you could suspect given to build like a space like this one there's way to many system
00:18:51
in that new to fit between two plans button phase of the of the slab and um
00:18:56
and the ceiling that was already way too many special collisions you could suspect
00:18:59
what was the numbers of design issue and special collision provided like this one
00:19:04
so we just trying to help not uh by given a catalogue of the numbers of problem but being more on the problem
00:19:09
solving and trying to to find a systemic way if you
00:19:12
inject dissolution here you may fix ten uh problems instead of one
00:19:17
uh there was a very formal process to document the design issues
00:19:21
but then we we develop also tool that uh we we call it i believe in what i see people
00:19:26
just really like it's a version and plus one that is a tracking look into the version on the server
00:19:31
and uh the balloon stern green only one another's like a
00:19:35
proven a a solution problem everything have been automatically um made
00:19:41
so this is the kind of hits map that show you the the flame so
00:19:45
the scale of the sphere give you how yeah how critical is the design issue
00:19:48
in a year cycle so we we turn most of the red balloons um uh to green
00:19:54
this is the sandwich i'm i'm talk uh was talking about so you have
00:19:57
the clapping envelop you have the steel structures and you have the him to be
00:20:01
um we've been running different set of corporation number one just steal and steel was
00:20:06
already like a highly complex iverson they does a remarkable job uh use into colour
00:20:11
uh this is between steel and concrete any p. um and then we introduce a also these
00:20:21
very simple concept of digital markup so would be just looking uh like a high fidelity want
00:20:25
to one looking into different a concept of assembly and fabrication putting them together before appointing the
00:20:31
uh the security pavilion as the markup scale one just to test that all those those presses
00:20:37
uh all the layout for the uh for the cassette and for the noise
00:20:43
so uh those also but it attention about how the disk and so novell was looking just for like
00:20:48
a sharp edge a construct which is simulation from a micro scale likable to learn different amounts and into
00:20:54
like a few days uh what could be assemble those about ten panels for by team to to to
00:21:01
clad i'm forty square meters so we had like ten teams as running ca to plug in the your time
00:21:06
uh negroes can just looking to make that statement understanding how to some like a small piece one after the other
00:21:12
um this is something that really i think what could just summarise what we're trying to do
00:21:16
just to eliminate the gap between the digital and the physical word having a kind of scale one
00:21:21
uh this is how we start and how we'd and so uh you
00:21:24
know john to see big as bill so we've been just injecting half round
00:21:28
and trying to ensure that would have been built is exactly what have been designed and you don't need just
00:21:32
all these like energy to three disk and and do all those like a very uh time consuming uh things
00:21:39
um this is what the inside is a museum so there was like a like a very interesting experience
00:21:45
a fantastic work by a jane team and by
00:21:47
other designers that this is the circular display uh
00:21:52
uh the zero line like uh everything is curved is not flat
00:21:56
surfaces really um uh interesting springs from how torn out from inside
00:22:00
this is a small example just to show you how the model
00:22:03
was been using extensively on that not even a database but only base
00:22:07
a shark need to be suspended but you know this like a then so
00:22:10
we'd understand a um it it need to to reach the primary not the secondary
00:22:15
uh we've been uh we haven't different videos that our project is in a curve screen so when you look
00:22:20
at it just to some optic and try to understand what's the optimum a location for the for the projector
00:22:26
this is the shark uh and just like us a few snapshots from the other finish building a from outdoor
00:22:41
but you how
Conference Program
Implementing a quality assurance program for sustainable buildings
Don Neff, LJP Construction Services, Irvine, United States
Oct. 28, 2019 · 4:34 p.m.
317 views
Performance of heat recovery using an Active Double Skin Façade
Michael Bulander & Kishore Patel, Harley Ellis Devereaux, Los Angeles, United States
Oct. 28, 2019 · 4:47 p.m.
468 views
A bio-inspired solution to raise the efficiency of kinetic façades
Steven Ware, Art & Build Architects, Paris, France
Oct. 28, 2019 · 5:36 p.m.
155 views
Concrete curtain walls - Status, benefits and visions
Stephan Giesser, Solidian, Germany
Oct. 29, 2019 · 8:32 a.m.
153 views
Advances in passive and active coatings technology for glass
Rory Back, NSG Pilkington, Lathom, United Kingdom
Oct. 29, 2019 · 9:59 a.m.
194 views
Why shading must be dynamic
Anders Hall, Somfy, Sweden
Oct. 29, 2019 · 10:46 a.m.
745 views
Glazing and shading that make the difference
Oskar Storm, Saint-Gobain Glass, Sweden
Oct. 29, 2019 · 11:01 a.m.
5344 views
Successful façade design by simulation
Sven Moosberger, EQUA Solutions, Switzerland
Oct. 29, 2019 · 11:17 a.m.
181 views
Digitally printed multi-colored PV for building integration
Dirk Hengevoss & Michaela Terwilliger, Resp: FHNW, Switzerland - Hochschule Luzern, Switzerland
Oct. 29, 2019 · 11:20 a.m.
Predictive controls for HVAC systems
Urs Grossenbacher, Pronoó AG, Givisiez, Switzerland
Oct. 29, 2019 · 11:32 a.m.
Holistic, multi-dimensional analysis of shading impacts on an all-glass office building
Jason Kirkpatrick, Interface Engineering, San Francisco, United States
Oct. 29, 2019 · 11:44 a.m.
Advanced daylight modeling of façade systems for energy and comfort analysis
David Geisler-Moroder, Bartenbach GmbH, Aldrans, Austria
Oct. 29, 2019 · 12:16 p.m.
New building glass envelope with Light, Color and Images
Ion Luh, Consullux Lighting Consultants/CEL, Canada
Oct. 29, 2019 · 2:49 p.m.
The Desert Rose - When poesy meets technology
Nassim Saoud, Trimble Consulting, Paris, France
Oct. 29, 2019 · 4:30 p.m.
372 views
Recommended talks
Une app sans map ? Vous n’y pensez pas !
Emanuele Gennai, ESRI Suisse
Oct. 31, 2017 · 10:34 a.m.
186 views
