Topological Queries and Analysis of School Buildings Based ...

Topological Queries and Analysis of School BuildingsBased on Building Information Modeling (BIM) UsingParametric Design Tools and Visual Programming toDevelop New Building Typologies

Jessica Bielski1, Christoph Langenhan2, Babara Weyand3,Markus Neuber4, Viktor Eisenstadt5, Klaus-Dieter Althoff61,2Chair of Architectural Informatics, Faculty of Architecture, Technical Univer-sity of Munich, Germany 3,4ALN - Architekturbüro Leinhäupl + Neuber GmbH,Germany 5,6German Research Center for Artificial Intelligence (DFKI), Germany1,2{jessica.bielski|langenhan}@tum.de3,4{babara.weyand|markus.neuber}@leinhaeupl-neuber.de 5,6{viktor.eisenstadt|klaus-dieter.althoff}@dfki.de

School buildings are currently one of the largest portions of planning andbuilding projects in Germany. In order to reflect the continuous developments inschool building construction with constantly changing spatial requirements, anapproach to analyse, derive and combine patterns of schools is proposed to adaptschool typologies accordingly. Therefore, the topology is analysed, concerninginterconnection methods, such as adjacency, accessibility, depth, and flow. Thegeometric analysis of e.g. room sizes or spatial proportions is enhanced byincluding grouping of rooms, estimated room clusters, or room shapes.Furthermore, text-matching is used to determine e.g. room program fulfilment, orassigning functional room descriptions to predefined room types, revealing hugedifferences of terms throughout time and architects. First results of the analysesshow a relevant correlation between spatial proportion and room types.

Keywords: school building typologies, building information modeling (BIM),artificial intelligence (AI), topology, spatial analysis, digital semantic model

INTRODUCTIONThe development of architecture is characterised bycontinuous change due to social, ecological andtechnological conditions. Contemporary and futurebuilding tasks take place against the background

of the change of these framework conditions. Thechange in formal language of architecture, becauseof industrialisation and composite building mate-rial, resulted in classical modernism, with representa-tives like Le Corbusier or the Bauhaus, as the starting

D2.T8.S1. THE COGNITIVE CITY (AI) - Volume 2 - eCAADe 38 | 279

point of the designmethodologymovement (Richter2010).

The ideas of classical modernism in architectureand novel scientific and computer-aided methodsled to the first generation of the design methodol-ogy movement in the 1950s and 1960s, e.g. repre-sented by Christopher Alexander or Niklas Luhmann.In the 1970s the second generation of the designmethodologymovement, representedbyHorst Ritteland others, did not view design procedurally as thefulfilmentof requirements, but rather as an individualprocess that could only be described incompletely(Richter 2010). In the 1980s, the digital approachesof case-based reasoning (CBR) influenced the digitalbuilding design approaches, which led to the rise ofa field of AI research in the building industry, calledcase-based design (CBD), in the 1990s. Even thoughat a current perspective the systemisationof complexcases in architecture is not sufficiently solved and isreferred to as data acquisition bottleneck. For the for-malisation of building information, object-orientedapproaches (Eastman 1999) of product data mod-elling have been transferred to the construction in-dustry in the 1990s. To remedy these shortcomings,Langenhan (2010, 2013, 2017) introduced the ap-proach of “semantic building fingerprints” that facil-itates spatial relationships and their digital process-ing based on enricheddigital semantic building data.The digital semantic fingerprint of buildings was in-troduced to describe themain features of the design,forming the basis for similarity assessment to dealwith ambiguities and complexities of architecture.

After major school building construction activi-ties at the beginning of the 20th century and in the1970s, Germany is now experiencing a third waveof school construction (Wenzel 2019). The renova-tion backlog of existing buildings and the currentpopulation growth requires school expansions andnew buildings for all age groups and levels of ed-ucation. The influx into conurbations is aggravat-ing the situation regionally. In addition to the num-ber of pupils and the necessary capacities, other use-specific changes demand structural quality and flex-

ibility of contemporary school buildings, e.g. for all-day schools, forwhich the buildings designed as half-day schools are insufficiently designed. Moreover,new pedagogical concepts for integration, inclusionand digitalisation require new floor plan typologies,while considering old or constant requirements, suchas large simultaneous circulation of people. Whilethese old and new requirements are partly contra-dictory, schools are currently being designed in elab-orate planning processes to develop new function-ing building typologies. Therefore, new functioningbuilding typologies have to be developed, while un-derstanding and classifying existing facilities.

PROBLEM STATEMENTThe amount of school building constructions, ren-ovations, and expansions has constantly increased(Freireiss and Commerell 2019). Throughout history,schools, being the educational space, are a mirrorof the economical essence and ideal of the currenteconomy (Schmidt and Schuster 2014) by shapingthe child into the most suitable citizen for the coun-try’s current society (Opp and Bauer, 2010). Increas-ing globalisation results in an international compe-tition for high quality education and intellectuals asnational assets (Freireiss and Commerell 2019). Fur-thermore schools, representing a point of congre-gation and integration, are conceptual entities forcultural, social and intellectual education, as well asphysical institutions, providing space for cultivatingand developing the urban context.

These educational facilities are the space, wherea child is consciously perceiving social experience, re-sponsibilities and possibilities within a society for thefirst time. This artificial and controlled ‘Micro Soci-ety’ (Schmidt and Schuster 2014) is separated fromthe exterior space to trial and learn, described asthe ‘Third space for social interaction’ by Opp andBrosch (Opp and Bauer 2014). As early as the 1950s,Reisinger and Schirmer express the need for schoolbuildings to follow the new pedagogical methods,declaring it the ‘second home’, (1955), verified morethan 60 years later by Djahanschah, Auer, and Nagler

280 | eCAADe 38 - D2.T8.S1. THE COGNITIVE CITY (AI) - Volume 2

(2018). By exceeding the studying space and pro-viding a suitable atmosphere for children, the schoolensures thewell-being througharchitecture psychol-ogy (Opp and Bauer 2010) and the four qualities ofresidency, the thermic, hygienic, visual and acoustic(Djahanschah, Auer, and Nagler 2018). In order tomaximise the capabilities and progress of the child,an efficient educational infrastructure and a versatile,but appropriate, environment must be introduced.The pedagogue Gerald Hüthers bases this complexambience on amultifaceted experience, and the con-frontation of the familiar and the unknown, stimulat-ing the mental and intellectual growth (Schmidt andSchuster 2014). Opp and Bauer (2010), and Freireissand Commerell (2019) go even further to declare the‘Room’ itself as the entity of the ‘Third teacher’ by util-ising the familiar space, derived from the pedagogi-cal concepts of Reggio. Overall, a school building isa physical institution, which serves as a frameworkto deliver the entity of a pedagogical concept to in-crease intellectual development, the social upbring-ing, and cultural education.

In the introduction of the ‘Grundrissfibel’ (eng.:Reference work for floor plans) for school buildingsHönig and Nashed state that the actual facilities of-ten outlive pedagogical concepts. Therefore, thesebuildings must be much more flexible and robustthan other typologies (2015). Reisinger and Schirmerstrongly agree to design, plan and build schools withthe future in mind (1955). Djahanschah, Auer, andNagler use the concept of ‘Kaizen’ as the basis fortheir designs. The Japanese concept suggests thatevery product or process can be improved (2018),resulting in a constant change of school buildings(Freireiss and Commerell 2019). It does not neces-sarily call for the demolition and replacement of oldfacilities, but alsomaintaining old, functioning build-ings, which only need minor renovations because oftheir long-time acceptance by society and users (Dja-hanschah, Auer, and Nagler 2018).

Due to these various reasons, such as historicalandpolitical context, social and cultural aspects, ped-agogical influences, and developments within archi-

tecture, the typologies of schools are very complexbecause of e.g. legal regulations and changing re-quirements on schools because of e.g. group teach-ing concepts, instead of frontal teaching by an au-thority figure. Schools are designed through an in-terdisciplinary process with careful consideration ofall aspects and users of the building, like teachers,students, personel, and the public. In order to cat-egorise and recognise patterns in a case-base, solelyessential information is to be extracted from the vastamount of data, meta-information decoded and pro-cessed, and its terminology conformed. This schooldata will be used to generate school variations basedon a semi-automatic approach steering a combinedsystem that uses case-based reasoning and deep-learning technologies to analyse and classify facili-ties according to the external influence, such as his-tory and politics. Further, it is a planning support sys-tem for schools to rapidly get new school typologiesfor constantly changing requirements based on bestpractice schools.

RELATEDWORKDuring the first half of 19th century the facility ofa school was an open-plan room within the privatehome of the village teacher, called ‘Einraumschule’,where groups of different ages were assigned appro-priate tasks (Djahanschah, Auer, and Nagler 2018)(Schmidt and Schuster 2014).

As educational institutes are a political demon-stration of cultural and intellectual capabilities (Dja-hanschah, Auer, and Nagler 2018), in the 19th cen-tury schools and educational concepts, such asfrontal teaching, were based on military methods.The ‘Gangschule’ (eng.: Corridor school) was built foroptimising this teachingmethod of large classes, ap-proximately 70 students, which aligned classroomsalong one large hallway within a single-depth build-ing (Djahanschah, Auer, and Nagler 2018) (see Figure1). Around 1900 the teachings ofMontessori were in-troduced, which contained an important componentfor social learning and assisting the child to learn foritself, still utilising an east-facing ‘Gangschule’ (Dja-


hanschah, Auer, and Nagler 2018).

Figure 1Schools inLichtenberg, Berlin.Schmidt, M. et al.,Schulgesellschaft:vom Dazwischenzum Lernraum,Future schoolbuildings, Berlin,Jovis, 2014, p.18.Especially after 1945, the construction of school

buildings rapidly increased (Djahanschah, Auer, andNagler 2018) and on a political level, German fed-eralism was reintroduced, including that each ofthe 16 federal states has its own ministry of edu-cation. Therefore, the different German states cre-ated different school systems and thus, differentschool building regulations, which are further spec-ified on a communal level. In order to distance them-selves from the former pedagogic methods, author-itarian politics and the architectural form languageof the ‘Gangschule’, pavilion schools were introduced(Djahanschah, Auer, and Nagler 2018). Newly con-structed buildings, which Reisinger and Schirmer, asemployees of theMinistry of Reconstruction of 1955,advocate, emphasize a lot on light, flexibility, expan-sion, and greenery for playing and movement, butstill separate the children by age and gender (1955).They implement these attributes and spaces into thetypical ‘Gangschule’ and finally, advise to use hous-ing technology on a reasonable level (Reisinger andSchirmer 1955).

The schools of the 1960s and 1970s became aspace for exchange, equality, information, studiesand retreat as the essence of democracy (Schmidtand Schuster 2014) (Djahanschah, Auer and Nagler2018). Due to the highest number of school en-rolment after the Second World War, school build-ings became much larger, often called school cen-tres, planned in the manner of a ‘Gangschule’ asa double-depth building or ‘Atriumschule’ (eng.:Atrium school), using light-weight construction (Bay-erische Fertigbau GmbH 1974). Later it became

known that those buildings often led to inferior airquality and even to the ‘Sick-building-syndrome’, dueto fixed glazing and even faulty fully ventilated build-ings (Djahanschah, Auer, and Nagler 2018).

The performance of Germany in the PISA rankingin 2000 motivated a reformation of the pedagogicalmethods and the definite inclusion of pedagogicalconcepts into school building design was prompted,like Reisinger and Schirmer had already advised 65years ago (1955). As buildings need to support theseconcepts, smaller bundled classrooms, separated byflexible walls are arranged around a central commonroom called ‘Marktplatz’ (eng.: Marketplace) (see Fig-ure 2). These study clusters or ‘Lernlandschaften’(eng.: Study landscapes), derived from the ‘Einraum-schule’ (Djahanschah, Auer, and Nagler 2018), sup-port the dynamic studying of students supportingother students, which stimulates intellectual growthand social skills (Opp and Bauer 2010). Because ofthese pedagogical concepts, mixed schools of differ-ent ages are preferred, as it improves the develop-ment of all the children andpromotionof talents, andsimplifies the transition from one grade into another(Freireiss and Commerell 2019). Further, the futureof educational methods suggest team teaching, orco-teaching, of at least two teachers, as well as thestudent becoming the teacher (Schmidt and Schus-ter 2014).

Figure 2First floor ofprimary school inKarlsfeld by ALNGmbH.

As largely both parents work full-time and studentscurrently remain more hours at school, space forstaying, retreat, and supervision is needed in con-temporary buildings (Djahanschah, Auer, and Nagler2018) (Kurz 2015). The circulation area is now used


as additional space or vacant classrooms are repur-posed during non-used times (Kurz 2015). Schmidtand Schuster call this the ‘In-between space’, thephysical representation of the pedagogical entity ofthe school (2014). The contemporary facilities areshaped by limited space, especially in metropoli-tan areas due to rising student numbers (Freireissand Commerell 2019) and declining space. Conse-quently, the area per student has increased. Dur-ing the 19th century it was about 0.9 m², between1945 and 2000 it increased to 2m² and finally, in the‘Lernlandschaften’ 4m² are provided per child (Dja-hanschah, Auer, andNagler 2018). Furthermore, Kurzsees the need for school buildings to open them-selves up to the public and offer space as a culturaland social convergence point and for further educa-tion for adults (2015). Conveying the notion of a cul-tural centre amidst a district, schools can also presentthemselves as a public space, by incorporating ser-vice centres, public playgrounds and communitycentres (Schmidt and Schuster 2014). Furthermore,these buildings can be used as landmarks within ar-eas for orientation and urban planning, due to theirform language and positioning (Djahanschah, AuerandNagler 2018). Therefore, Freireiss and Commerellcall the school an ‘urban component’ (2019).

The experimental ‘Open school’ in Uto in Japanis a possible future school typology (see Figure 3). L-shapedwalls define anopen spaceof a roomwithno-madic classes, which are changing location for everyclass. It is supported by the team teaching methodand the ability of the rooms to open up the facadeand include the exterior as further space in summer.

Figure 3‘Open School‘ inUto, Japan.Schmidt, M. andSchuster, R.,Schulgesellschaft:vom Dazwischenzum Lernraum,Future schoolbuildings, Berlin,Jovis, 2014, p. 93.

APPROACHDue to this high amount of factors and require-ments, schools include a complex topology and highamount of room types. Even though thedesign focusof the contemporary twelve schools of ‘Future schoolbuildings’ (Freireiss and Commerell 2019) is differ-ent, the room program is quite similar. These build-ings are commonly two storeys high in the country-side, while within the metropolitan area and its re-sulting lack of space, they tend to be higher, risingup to a maximum of five floors. The different facil-ities focus on the pedagogical integration of multi-ple school typeswith classrooms, study rooms, grouprooms, laboratories and space for music, art andtheatrical teachings, while providing separate roomsfor personnel and teachers, administration, housingtechnology, storage, meeting rooms, sports facilities,and a large meeting hall, called ‘Aula’. A cafeteria,groupedwith a kitchenandpreparationarea, is incor-porated, as well as one or more multipurpose rooms,storage spaces, relaxation areas, playgrounds and afunction room for events, such as theatrical plays orconcerts as a point of congregation for social and cul-tural interaction through recurring events. Depend-ing on the curriculum, the room program needs toprovide specialised spaces like workshops or roomsequipped with advanced technology. The adjoininggreenery, an area for playing and movement, is of-ten used as a ‘Freiluftklassenzimmer’ (eng.: Outdoorclassroom)within anassigned space, sometimeswithexperimental gardens, as well as a public space.

To transform spatial configurations of schoolslike topological, semantic andgeometric informationin a computer readable way to serve as a case-basefor the generation of variations, using deep-learning,the necessary information is stored in the architec-turalGraphML (aGraphML) format (Langenhan2017),derived from the fluent graph format of GraphML,based on the XML format. It is used to depict adja-cency, accessibility, depth, flow and analysis of theroom program to the user in Rhino3D and Grasshop-per3D in a readable visualisation. The Dolphin plu-gin (Langenhan 2015) for Grasshopper3D, provides a


component to translate into said file format. In thefollowing, the workflow through the different soft-ware programs is described, as well as the differentsteps to convert the extracted information of an In-dustry FoundationClasses (IFC) file into an aGraphMLformat to analyse school buildings and generate vari-ations as part of our future work. Furthermore, theresults of the variant generation is to be used in Au-todesk Revit in later stages.

This workflow offers a software overview, pictur-ing the different programs and plugins, which arecurrently available, to translate an IFC file formatinto an aGraphML file format (see Figure 4). For thisproject REVIT 2020 fromAutodeskwas used. The plu-gin Rhino.Inside for REVIT 2020, developed by Mc-Neel, is based on the Rhino3D 7 Work-In-Progress(WIP). It allows extracting and working in real-timewith REVIT elements in a separate Rhino window,

Figure 4Overview andworkflow throughsoftware programs.


open in REVIT. Working in Rhino3D offers to workwith the parametric design tool Grasshopper for vi-sual programming. Various plugins needed to be in-stalled in Grasshopper, such as Pufferfish, TT Tool-box, Kangaroo2, PandaandEleFront, to transform theavailable geometries into readable shape language.Further, Rhino.Inside for Revit offers custom compo-nents, such as the Category.Picker, Elements.Filter,Category.Identity and Category.Elements. Used

correctly, the chosen category elements show asghosted objects in the Rhino window, as well as theREVIT main window, proving the real-time capabili-ties of Rhino.Inside. The plugin EleFront is used toassign attributes to the to-be-baked objects. Fol-lowing, the Grasshopper plugin Dolphin, active inGrasshopper, is able to translate these shapes intothe aGraphML format internally. This format thenwas printed as an XML file into the desired path.

Figure 5Workflow of steps.


The first work step, as depicted in Figure 5, is ex-tracting the relevant data, using the custom compo-nents of Rhino.Inside. The following tasks are simul-taneously taken, but apart from each other. In or-der to create separated storeys as 2DObjects,meshesneed to be transformed into Boundary Representa-tions (Breps) and their base surfaces, grouped by theZ value and transformed into circles for readabilityby the Dolphin plugin. Afterwards, by using the ap-propriate list for the desired storey level, the cho-sen objects are isolated. In order to compare differ-ent floor plans of schools, the terminology of roomnames must be aligned. The research on the twelvebuildings of ‘Future school buildings’ (Freireiss andCommerell, 2019) is used as reference to determinea dictionary. The different room names, assigned bydifferent people, organisation, and of different times,must be interpreted by an architect and individuallytranslated. At first the extracted data, represented ina list, needs to be categorised and then conformedunder an overlying term. Text-matching is imple-mented, so keywords or partial words can be usedto change the ‘Room names’ appropriately. Finally,the different categories are baked back into Rhino3D,merging the 2D objects and their conformed seman-tics, in order for the plugin Dolphin to read and con-vert it into aGraphML file. As a result a rough topo-logical analysis was performed to identify structuralpatterns for school building as a basis to develop au-tomatic machine learning approaches. For example,differentiating by names and ranking the sizes, aswell as the amount of edges, using Gephi.

EVALUATION AND RESULTSThe described workflow and process in Grasshopperhas been applied to IFC files of three different schoolbuildings of the architect’s office of ALN, Architek-turbüro Leinhäupl + Neuber GmbH. Due to the test-ing with these floor plans, necessary conditions weredepicted, improvements in the visual programminghave been made, and finally, first pattern recogni-tionswere conducted, using the createdvisual graph.

The Dolphin plugin has various restrictions, suchas the representative room shapes of circles and thatthe interconnection between two circles cannot bemore than one. Further, one circle cannot have morethan one room descriptions. During testing, over-lays occurred, as the interconnected points and roomnames are the centreof the circles. Improvements us-ing the available floor plans, led to the developmentof an automated point correction.

The first tests showed satisfying and encourag-ing results (see Figure 6). After the automatic point-correction, the aGraphML files can be read, as well asthe visual graph is well readable. The first pattern,recognised by the architects, is a correlation betweenspace and size, such as the circulation area, which isvery large and often connects or is itself a multipur-pose area. As most of the floor plans include studyclusters, it becomes clear, that the rooms open to stu-dents, are all quite similar in size and shape, but as thechildren can occupy all of them, the 4m² per studentare achieved. Further, the sequenced succession ofrooms, starting at the ‘Marktplatz’ and ending in theclassroom is evolving from a large common roomalternating to a classroom through a passage-likesmall room, called ‘inclusions’, or a supervision room,which can be accessed from both adjoining class-rooms. Gephi revealed that a classroom is linkedwithat least oneother classroom. Group rooms are alwaysconnected to at least one ‘Unterrichtsraum’ (eng.:Teaching room), which is different from the class-room. Simultaneously, group rooms take up abouthalf the area size of a classroom or equally sized ‘Un-terrichtsraum’, thus providing the combined amountof space of ‘Unterrichtsraum’ and group roomduringproject works, maximising the available square me-tre per student. Further, the ground floor of schoolbuildings with more than two storeys is used solelyfor administrative and public rooms, such as library,cafeteria or first aid.


Figure 6Developmentthrough theprocess.

FUTUREWORKFor further work, the Dolphin plugin is currently be-ing improved. After the first developments, it recog-nises polygon shapes as rooms, simplifying the entireprocess of creating ‘Wall’ and ‘Door’ line representa-tion and offering the ability to write the actual roomedges, simultaneously solving the Boolean edge at-

tribute of ‘enclosed room’. Further, a new library forroom terminology was created, so the ‘Room names’are not converted, but ‘Room types’ can be recog-nised by ‘Room name’ and appropriately written. Fi-nally, an edge representation for windows was cre-ated, as well as a Boolean option for ‘Windows ex-ist’ edge attribute. The latest developments evolvearound generic ‘Exterior’ Nodes and ‘Entrance’ edge


types. The solution for this includes both improvingthe Grasshopper components of Dolphin, as well asvisual programming using Grasshopper3D.

Finally, after more testing on a larger scale offloor plans, the pre-processed data is to be intro-duced to an artificial neural network to generatenew architectural floor plans, using Deep GenerativeModelling. An algorithmic setup issued to build Ar-tificial Neural Networks to generate new data, basedon the pre-processed data provided to the model, assuggested inArora et al. (2020). This data canbeusede.g. with a tensor data structure ‘relation map’, wepropose in Eisenstadt et al. (2020) for intelligent vari-ant generation, according to predefined design re-quirements and finally, an auto-completion of build-ing designs.

ACKNOWLEDGEMENTSThis research is part of the research project metis-II,funded by the German Research Foundation (DFG).The test floor plans are provided by the architec-ture office ALN, Architekturbüro Leinhäupl + NeuberGmbH.

REFERENCESArora, H, Langenhan, C, Petzold, F, Eisenstadt, V and Al-

thoff, K-D 2020 ’Autocomplete designs by suggest-ingbest practicedesign solutions indigital semanticbuilding models (BIM) by combining AI approachesof case-based design and deep learning’, Proceed-ings of ECPPM 2020, Moscow (accepted)

Djahanschah, S, Auer, T and Nagler, F (eds) 2017,Zukunftsfähiger Schulbau, Detail Business Informa-tion GmbH, München

Eastman, CM 1999, Building Product Models: ComputerEnvironments Supporting Design and Construction,CRC Press, Boca Raton, Florida

Eisenstadt, V, Langenhan, C, Althoff, K-D and Dengel, A2020 ’Improved and Visually Enhanced Case-BasedRetrieval of Room Configurations for Assistance inArchitectural Design Education’, Proceedings of IC-CBR 2020, Salamanca

Freireiss, K and Commerell, HJ 2019 ’Zukunft Schulbau:Europäische Beispiele zeitgemäßer Schularchitek-tur [exhibition catalogue]’, Aedes Architekturforum,Berlin

Bayerische Fertigbau GmbH, - 1974, uniplan Schulbau,Lizenz imbau, Garching b. München

Hönig, R and Nashed, L 2015, ’Einleitung’, in Hönig,R (eds) 2015, Grundrissfibel: Schulbauten, EditionHochparterre, Zürich, p. 6,7

Kurz, D 2015, ’Entwicklungen im Schulbau’, in Hönig,R (eds) 2015, Grundrissfibel: Schulbauten, EditionHochparterre, Zürich, pp. 8-19

Langenhan, C 2017, Datenmanagement in der Ar-chitektur. Untersuchung zur Organisation vonEntwurfsinformationen in IT-Infastrukturen undNutzungsmöglichkeiten in wissensbasierten Syste-men., Ph.D. Thesis, Technische Universität München

Langenhan, C and Petzold, F 2010, ’The Fingerprint ofArchitecture: Sketch-Based Design Methods of Re-searching Building Layouts through the SemanticFingerprint of Floor Plans’, JournalofArchitectureandModern Information Technologies, 4, pp. 4-13

Langenhan, C and Petzold, F 2015 ’BEYOND THE BUBBLE- Computer-aided topological analysis andparamet-ric design of room configurations in university edu-cation.’, Proceedings of eCAADe 2015, Vienna

Langenhan, C, Weber, M, Liwicki, M, Petzold, F and Den-gel, A 2013, ’Graph-based retrieval of building in-forrmation models for supporting the early designstages.’, Advanced Engineering Informatics, 27(4), pp.413-426

Opp, G and Bauer, A (eds) 2010, Lebensraum Schule:Raumkonzepte planen, gestalten, entwickeln,Fraunhofer-IRB Verlag, Stuttgart

Reisinger, MLR and Schirmer, F 1955, Richtlinien für denSchulbau, August Bagel Verlag, Düsseldorf

Richter, K 2010, Augmenting Designers’ Memory: CaseBased Reasoning in der Architektur, Ph.D. Thesis,Bauhaus-Universität Weimar

Schmidt, M and Schuster, R (eds) 2014, Schulgesellschaft:vomDazwischen zum Lernraum; 30 Schulgebäude imVergleich, Jovis, Berlin

[1] https://www.competitionline.com/de/news/markt/monitor2019-wundertuete-schulbau-2136.html


Topological Queries and Analysis of School Buildings Based ...

Documents

Transcript of Topological Queries and Analysis of School Buildings Based ...