Skip to content Skip to sidebar Skip to footer

Source: batiactu as of 26-03-2020

Exhibition. The   Arsenal Pavilion  offers,  in  this confined period,   a  virtual tour  of  the exhibition “Artificial Intelligence artificielle  and Architecture”, inaugurated on  February 27. février From  the  origins of modularity  in the  années twenties,to the incredible possibilities   offered by the  programs programmes  in  development  at Harvard, Berkeley  or  MIT  today..

Artificial intelligence (AI) has already  invested en  many  industries and moyens   disciplines   by    empowering  them  to meet en  previously unprecedented challenges    and  making too complex operations  possible.  Its  use  in the  field  of  Architecture  is still in its  infancy,  but  the first  results  obtained  are  promising  and the prospects  numerous.” It was  this   initial  premise that led Stanislas  Chaillou, an architect  and  researcher  at Harvard, to  mount  the IA  and Architecture exhibition at the Arsenal Pavilion  in Paris. The  HYPERLINK “” \t “_blank”   as part of the architecture at programme   home..

It all begins,  according to the exhibition,at the beginning of the  twentieth  century, with studies on modularity.  The  starting point for the  systematization of architectural design, the modulargrid,   theorized  at the Bauhaus by Walter Gropius  as early as  1920,  carries  the hope  of technical simplicity  and the  promise  of economic architecture.  It will be  applied  to the  Dymaxion House (1929-1946), which  takes  modular housing  to  the  extreme, démonstration and  is  “the first  proof of the concept  for the industry”.”.  This  standardization  was later developed  by Le Corbusier’s Modulor  (1945), and  even influenced urban planning at the turn of the   sixties, when projects  such as Archigram’s  Plug-in City   (   theoretical project, 1964)  aim  to  create   fully  modular cities. .

Modularity, computer-aided   design, setting  and artificial intelligence.  A century of technological advances    in  architecture are presented to  us by the  Arsenal Pavilion. .

Computer-assisted  design (CAO), starting in  the sixties,  will  allow,by the  new computing power now  accessible, “the  systematization of architectural design, based on  strict rules  “. Christopher Alexander, an architect  himself  and a professor at California University (Berkeley),  establishes  one of the key  principles  of computer  design:    the “Object-OrientedProgramming”. The  principles  he  defines  in  his publications are,  even today,  the  essential foundations  of  computer programming.  In  his  footsteps,  of ” entire  generation of computer scientists  and architects  is creating a new field of research.”  

Zaha Hadid, pioneer  of  seterism

The third leg  of the trip  proposed  by the   Arsenal Pavilion:     the setting..  It  appears  in the  1990 svingtand  avoids   the repetitive tasks  imposed  by CAD applications, while  better  mastering complex forms.   Thanks  to  this design device,  “eachchaque  task  is  streamlined  into a set of simple rules,   constituting  a  procedure..  This  procedure  can  be  encoded  in the  program  by  the architect   in order  to  automate  an  execution,  previously  manual  and  tedious,”says  the  curator..  But,    while  en  allowing it    to encode  a  given procedure, “the  system  allows it  to  isolate  immediately  the  key parameters that affect the  result..  The architect  is  then  able  to  vary  his  own  parameters  in  order to   produce  different  scenarios:    shapes  or  options  generated  instantly by the simple modification of  the  parameters previously  defined.””.

Zaha Hadid(who died  in  2016)  is at the forefront of  this  process. Iraqi architect  and  mathematician    trained  in the   United  biais elle Kingdom, she combines  these  two  areas through parametric design.  “His work is usually the result  of  rules,  encoded  directly  in the  program,  allowing  an   unprecedented level  of  control over the  geometry  of  buildings..  Each  architectural  decision  results  in  a  given setting  of  parameters,which  in  turn  gives a  specific  shape to the  building, “explains  the exhibition .  Sketchpad, Pro/Ingeneer,Grasshopper,  in the 2000s,  and  ,  BIM’s  BIM main tool  today,  are  the  tools  that  have  accompanied  this  rise  to  power.  become  unavoidable, of the parameterism  in the architecture..

Artificial intelligence and  its  opportunities

Artificial en intelligence is finally    coming.  It  represents  a new wave of technology that”complements  the  practice  of the architect by assisting him from the  point of  view  of  architectural expertise  and  enriches  his expression.   Today,the results  of research,  academic  and  private,show  the first results  of  this  evolution. The techniques of  ai  known as    ‘generative’-  that is, able to  create shapes,not  just analyze  them  –  are  recent..  For  the past three years,  they  have opened up  new fields of experimentation.””.

The Arsenal Pavilion   explores the current scales    of experimentation:    plans,  elevations,structures and perspectives, in  which  AI  could  already  make  a  contribution,  real  or  speculative. Stanislas  Chaillou, curator of exhibition,made the work of  his  thesis at Harvard available  at   the Arsenal Pavilion. .  Other leading researchers  on  the subject, the Berkeley Artificial Intelligence Research (BAIR) Lab,  or the Digital Structures Lab  at  MIT (Massachusetts Institute of Technology),  have  lent  their  work.

Decisive advance

The “adverse generative   neural networks” are one of the  most  promising fields of  AI, and “conceptually  decisive advance”. “.  Explanation:  combining  two   models,the  generator and the  discriminator,  these  neural networks neurones  “proceed from  an  analogy  similar to the relationship between the student  and the teacher”: the  generator  (student)  seeks  to  generate  images   resembling that  of a  user-defined database.  The  discriminator (theteacher)  gives the generator  a  note  for  each new image  generated  by  the  generator. This  note  judges  the likelihood of the image  created..  En  Depending on   this   result,the  generator will adapt  to  improve  its  performance, just  as  a  student would work  to  get better grades.”  “.


Show CommentsClose Comments

Leave a comment

News ORS © 2020. All Rights Reserved.