A Grammar-based Procedural Design Guideline Visualization Diagram for the Development of SVA Masdar

Abstract. Nowadays, a large set of involved planning parties are heavily demanded with the definition of holistic in kind requirement specifications for urban planning sites – so called future cities. However, the resulting amount of specifications
of 8
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  eCAADe 28 833 -City Modelling A Grammar-based Procedural Design GuidelineVisualization Diagram for the Development of SVAMasdar  Jan Halatsch 1  , Thomas Caro 2  , Bruno Moser  3  , Gerhard Schmitt  41,4  ETH Zurich, Switzerland, 2, Switzerland, 3  Foster+Partners, United Kingdom 1, 2, 3 1,, 3, 4 Abstract. Nowadays, a large set of involved planning parties are heavilydemanded with the denition of holistic in kind requirement specications for urban planning sites – so called future cities. However, the resulting amount of specications for a specic building project poses a great challenge todesigners and planners especially when it comes to include this informationinto their design proposals for a sustainable urban development. Thesedesign performance criteria are traditionally expressed in textual and numerical planning guidelines and which are making it difcult to establish acomprehensive and holistic view onto the domain itself. Therefore we present in this paper a design guide visualization method to overcome this situation for the evaluation of design specication and urban layouts in a qualitative and quantitative manner. Keywords. Sustainable urban patterns; shape grammars; design evaluation;urban planning; design guide translation.  Figure 1 A design guide-line visu-alization diagram can helpto communicate complexinterdependencies in urban planning.  834 eCAADe 28 -City Modelling Sustainable urban design rules have tobe human readable Our method stands in to reduce the inherited com-plexity o such requirement specifcations with avisual and interactive parametric ‘diagram’ that isbased on a generic city system model (Figure 1). Ituses grammar-based shape patterns, eature pat-terns and conditional patterns to describe a large seto regulations that results into a n-dimensional de-sign guide visualization diagram. The method incor-porates the context sensitive and stochastic distribu-tion o geometric and meta-inormal design eaturesand combines it with grammar-based geometrygeneration and conditional property verifcation. The grammar-based implementation is procedur-ally linked and thereore operating seamlessly romthe regional planning scale up to the architecturalbuilding scale. Further it allows the user to set-up acity inormation model environment that includesgeometric scopes and entities, conditional behav-iors and city properties. The resulting model can beused to interpret the complex nature o sustainablecity developments and to communicate more e-fciently system specifc eatures to stakeholders o a development. We exempliy our method with thevisualization o Foster+Partners planning guidelinesor the development o the Swiss Village in MasdarCity, Abu Dhabi (SVA) [1], which presents by itsel anoutstanding prototype or carbon-dioxide neutraluture city. The visualization o the SVA serves as acollaborative platorm and interactive visualizationramework or the SVA’s architects and potential in-vestors. It communicates visually the key propertieso the underlying guidelines and enables the user atthe same time to reect local design pattern typolo-gies in cohesion with benchmark criteria or examplematerial masses that are planned to be brought intoa certain region o the SVA and the actual GFA thathad been achieved. Architects and planners caneasily adapt the parameters o the model and iteratethrough dierent kind o versions o the model toachieve or example lower energy consumption dueto a rearranged placement o solar shades in ronto buildings inside Masdar. In this way the model byitsel is meant to interace the design guidelines withthe fnal design proposal handed in by the architects.With such an interactive model they can understandthe relationships between critical model proper-ties and use these insights or a verifcation o theirown design implementation or example i it is likelyto ollow the rules o Foster+Partners as the initialideas or the Masdar planning site. Future work willas well deal with linking the underlying proceduralcity model with the eco-system-service inormationrom the landscape and environmental planners aswell as to combine both with an economic model orthe case study in Masdar. Overview  The paper is structured as ollows: The next sectionpresents the related work in the feld. Section 2.1paraphrases the use o the design pattern paradigmin procedural urban modeling. The ollowing section2.2 explains the process o translating textual andnumerical data into an interactive procedural model.Section 2.3 presents the interactive design guide-line visualization diagram. Selected case studies arepresented in Section 3. Finally, Section 4 addressesconclusion and uture work and with Section 5 ac-knowledgements are made. Related work   The system reers to the ones by Müller et al. (2006)and Halatsch et al. (2008) with urban planning rulesets and landscape patterns, which can be used orpre-visualization, master planning, guided designvariation and or digital content creation purposes o the entertainment industries. The rules and patternsare associated with architectural attributes. Gener-ated geometries ollow basic architectural normsaccording to GIS inormation. Beirão et al. (2008) al-ready introduced the structure o a shape grammarbased supporting tool or urban design, comprisingan urban pattern ormulation model, a design gen-eration model, and an evaluation model.  eCAADe 28 835 -City Modelling In Halatsch et al. (2009) the connections be-tween urban models and grammar-based proce-dural generation o urban environments and theirrelevance to urban planning have been introduced. The computational tool or the analysis and gen-eration o our generic city model is based on shapegrammars (Stiny, 2006). Shape Grammars have beenimplemented in the past in the analysis o severalhistorical examples, such as the Palladian Villas (Stinyand Mitchell, 1978). The technical characteristics aredirectly derived rom an attributed shape grammarcalled CGA Shape, which is suited or applications incomputer graphics. It was introduced by Müller et al.(2006) and had been extended by Ulmer et al. (2007),Halatsch et al. (2008) with urban planning rule setsand landscape patterns, which can be used or pre-visualization, master planning, guided design varia-tion and or digital content creation purposes o theentertainment industries. The rules and patterns areassociated with architectural attributes. Generatedgeometries ollow basic architectural norms accord-ing to GIS inormation.With this ramework, a wide range o architec-tural designs and urban layouts can be encoded anddetailed 3D models can be generated automatically.Shape grammar rules that create certain architec-tural confgurations can be grouped into collectionswithin libraries. Such libraries exist in literary ormand are described as patterns. The use o patternsin architecture is ounded on Alexander et al (1977). An integrated city evaluation model  The implementation o the generic city model isdone inside a joint research project with ProceduralInc. [2] by extending their commercially availableCityEngine application. We use CityEngine’s CGAshape grammar approach and CityEngine’s GIS inter-connectivity to access and exchange data. As inputsdesign data and unctional data (e.g. social statistics,or approximated energy consumption) are used,which are then processed by the system. The out-put results in a parametric city model by triggeringspecifc design grammar rules during the system’sderivation process. The process is steered by socalled patterns, which are imported rom standardGIS, CAD and image processing systems or modeledinside the system. In more detail, the output o thegrammar derivation process embodies architecturalgeometry and, or the visualization o spatial relation-ships between the entities o a city. A pattern-based modeling approach Usually, building laws or best practice examples (asa part o empiric planning knowledge) and personalpreerences by planners need to be explicitly andimplicitly incorporated into an urban design and atdierent scales, e.g. rom regional to building level. The underlying set o data is more or less only avail-able in the orm o textbook or table sheet defni-tions. Furthermore, an integrated computer-basedmodel, which is highly interactive, does not yet exist.Whereas more and more data is recently integratedby municipalities into GIS databases, which oer aunique opportunity to evaluate existing structuresor to predict urban scenarios.Our generic system steps in to create multi-dimensional models that can be used to evaluateaspects o urban perormance and it can be usedto create visual representations or experts and lay-people at the same time. The underlying systemmodel is implemented in the orm o a proceduralarbitration o dierent urban scales that can beseamlessly traversed rom regional to building level. This system overcomes the traditional separation o urban data into GIS-based large-scale models andmodern CAAD systems’ small-scale building inor-mation models. Even more important, the systemopens the possibility to have interactive parametricurban models at hand, which can semi-automatical-ly generate 3D visualizations. These urban modelsare not restricted to contain 3D data representa-tions or one given level o detail. They can easilyand instantly modifed and adjusted seamlessly. Forexample, energy analysis on an urban scale can beperormed and based on the results new regulations  836 eCAADe 28 -City Modelling can be ormulated and incorporated into passiveenergy eciency guidelines or urban sites. The up-dated regulations can easily be compared to exist-ing defnitions and then evaluated with having, e.g.,(a) concrete numbers, (b) a design catalog and (c) a3D representation at hand. Besides that, authoringsotware packages or procedural models are moreand more integrated with recent GIS-sotware. Inthe ollowing we are exempliying the process andthe implementation with Procedural’s CityEngine[2], a sophisticated procedural modeling ramework,which can be tightly integrated to ESRI’s ArcGIS [3]and exchange data with recent CAAD tools such asAutodesk Revit [4] and analysis tools such as Au-todesk Ecotect [4]. The visual and numeric eedback can be used to validate the procedural model or torefne and rework the model until each perormancecondition is ulflled. That leads to the interestingquestions how the text- or number-based inorma-tion rom regulations etc. may be translated andintegrated into a procedural model. Thereore, werecently discussed the use o design patterns ormaster planning (Ulmer et al. 2007, Halatsch et al.2008, and Wissen Hayek et al. 2010). The goal is thata planner may structure an urban model into urbansystems and subsystems. Each system can be de-scribed by design patterns, which represent statisti-cal unctions and values, geometric confgurations,conditional decision schemes that interact with andattached property values. So, what are patterns?And how they are used? We dierentiate into ormeecting and statistical patterns. E.g., grammar-based shape patterns are the geometric reactiono unctional or regulative constraints (economical,ecological, sociological). On the other hand, statis-tical patterns describe value distributions across acertain area, such as a targeted population densitydistribution steered by building height regulations.Descriptive values such as material names and habi-tat species are used to steer conditional patterns aswell as to report the fnal perormance o the gen-erated urban model. We use a conditional propertyverifcation to determine i a value is valid or a givensituation. Booleans comparisons between input val-ues are used to traverse through decision trees tofnd the most tolerable solution.In more detail we are structuring patterns intothe ollowing catagories:1. Grammar-based “shape modiying patterns”:Pure mass unctions such as extrusion, block types (perimeter, l-/U-/H-/L-shapes), genericbuilding and plot subdivision that are most dom-inant and regulative. Generic allocation o zonesand the global defnition that are expressed bylaw regulations e.g. maximum building heights,land use mix, solar envelopes or statistical valuedistribution such as population density.2. “Feature patterns” cover architectural aestheticsand unctional measures on a more subordinatelevel. • “Architectural design eature pattern”: De-sign o building mass confgurations andproportion o building design elements, inconormity, reaction or integration with “De-sign perormance eature pattern” • “Design perormance eature pattern”: Ful-fllment o unctional requirements, model-ing o ecological constraints such as (a) habi-tat demands, natural shading unctionality,(b) economic constraints such as u-values(energy transmission) per costs per sqm incomparison with grey CO2 stored inside theparticular material, or - on a more regionallevel - distance to work, education and sup-ply, (c) social actors such as accessibility dis-tributions to daily amenities and work, visi-bility and hal-public interaction on buildingscale3. Conditional patterns incorporate a behaviormodel that interact with context sensitive in-ormation (e.g. given ecologic, economic andsocial properties that have been defned in 1.and 2. and are integrated into the proceduralurban model). In particular conditional patternsare used to select appropriate eatures patterns,which will hold the targeted perormance crite-  eCAADe 28 837 -City Modelling ria or a defned eco-system-services. Statisticalunctions can be used to provide value rangesthat will select predefned eature patterns. Landuse, maximum building envelope, building eras,acade typologies or wall materials can be easilydefned and automatically applied to the urbanmodel. Especially in situations where only moreor less loose defnitions exist these values canstill be considered as inputs. Thereore the valueranges and the intended reactions have to beinitially defned in a conditional pattern.System elements are connected via GIS databaseintegration. The GIS database serves as a containersystem to access data or modeling, evaluation andthe generation o multi-dimensional spatial datasets. Results rom the generation can then be back-looped into the GIS database. The data integrationworks seamlessly. Several level-o-details (LODs) di-erentiations are perormed to distinguish betweendierent uses o the model. Block-wise extrusions(LOD 1) are used to evaluate large-scale models.Higher resolution LODs provide or instance moreaccurate material property defnitions and geomet-ric appearances as well as a higher granularity orthe reporting o building and other elements. Thisunctionality is important since on each scale di-erent evaluation techniques have to be applied orthe urban analysis o a modeled area. Each scale canprovide eedback and thereore interact with otherscales. The interaction between the interconnectedprocedural entities inside the model can be used tostudy interdependencies between properties andspecial ocus measures. The resulting inormationabout the interactions and interdependencies are in-tegrated into a standard GIS database. From the da-tabase itsel the inormation o the model, the analy-sis and the evaluation reports can be easily accessedwithout special knowledge o our system. Further-more, the stored inormation and results can easilybe distributed to other sotware packages, e.g., orCFD simulations (computational uid dynamics) or3D real-time visualization environments as well as torapid prototyping applications. Translation process: from textual descriptionto city model  The process o transorming abstract regulationsinto a more understandable and interactive dia-grams is structured as ollows. First or each scale thetop 10 measures have to be identifed. These mea-sures should then be incorporated into an orderedlist. The measures may be statistical values such asthe population density at a given point or a geomet-ric attribute, e.g., the overhang o a building to pro-vide shading on the sidewalk or pedestrians. Basedon that order the analyst has to defne, e.g. on block scale, a mass model that will be detailed enough toulfll the initially selected condition. The mass modelmay be described by a hand sketch (Figure 2). Next  Figure 2 From left to right: At rst an isometric gure has to bedrafted, which incorporate all necessary patterns. Second,the inherited hierarchy insidethe sketch has to be proce-durally implemented into several level of details (LOD). Finally, the initial compu-tational massing is rened,transformed and applied toan urban context by the useof feature and conditional  patterns.
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!