Human Behaviour Simulation Using Space Syntax Methods

Architecture and Urban Planning 2020, Vol. 16, Issue 1, pp. 84–92 https://doi.org/10.2478/aup-2020-0013 https://content.sciendo.com Online ISSN 2255-8764 Human Behaviour Simulation Using Space Syntax Methods Milda Sutkaitytė*, Kaunas University of Technology, Kaunas, Lithuania Abstract – City is a multi-layered structure of social, cultural, and economic aspects and their relationship through the physical space. Recognition of some patterns in those relationships is the essence for defining fragmentations in urban fabric and suggesting solutions on how those fragmentations could be solved. The article analyses how different space syntax methods can be used to find patterns in the chosen urban environment. Space syntax allows to find urban relationships between physical environment and human behavior. Space syntax suggests a few different approaches on how these relationships could be simulated: Segment Analysis perceives environment as a network of paths or streets, visibility graph analysis concentrates on inter-visual relationships, while agent-based analysis uses simple artificial intelligence for modeling movement in open space. Consequentially, the aim of this research is to find out what human behaviour aspects each of these space syntax methods are able to simulate. Keywords – agent-based analysis, segment analysis, Soviet micro-districts, social control, Space syntax, urban network, visibility graph analysis. I ntroductIon Urban environment is a complex system of social, economic and spatial relationships. All those relationships are better understood if certain patterns are found, most importantly – patterns in physical environment, human behaviour and connection between them. Furthermore, recognition of those patterns helps not only to understand urban environment overall, but also to find fragmentations in urban fabric and to suggest solutions on how those fragmentations could be solved. Space syntax is a group of methods, based on simple mathematical rules and graph theory, helping to find patterns in given environment. Nonetheless, space syntax is a bottom-up simulation able to recognize smallest elements’ behaviour and to visually present the aggregated view to the observer. These features of space syntax are particularly helpful in finding urban relationships discussed above. On the other hand, space syntax suggests a few different approaches on how these relationships could be simulated: segment analysis perceives environment as a network of paths or streets, visibility graph analysis concentrates on inter-visual relationships, while agent-based analysis uses simple artificial intelligence for modeling movement in open space. Consequentially, the aim of this research is to find out what human behaviour aspects each of these space syntax methods are able to simulate. A challenging urban environment was chosen as a test site for this research, as it has a variety of different problems to solve, but the findings obtained could be useful to improve numerous other locations. Soviet micro-districts were chosen for this task as they both have a lot of unsolved urban issues, such as mono-functionality and segregation from other city parts, absence of urban private zones, ignorance to human scale and need for identity, etc., and make up a large proportion of overall living environment in post-Soviet countries [1]. Nonetheless, Lithuanian Soviet micro-districts, as well as majority of districts of the same period in other Eastern Europe countries, are particularly sensitive to destructive changes, as the proportion of inhabitants there is high and extreme means would not simply be acceptable, comfortable and profitable for the residents and other stakeholders [2], [1]. Accurate identification of location, problem and means become of main importance when solving issues there, and Space Syntax supports these actions. I. M ethodology Methodology of the research comprises three stages: territory selection, conducting sociological survey, and space syntax analysis. All of these stages have their separate goals and their synergy ensures comprehensive results – useful globally not less than locally. The goals of the stages are as follows: 1. Territory selection – to select a limited number of Soviet micro-districts by analysis of development chronology, morphotype, acknowledgement, fractal index and continuity of experiment territory. The selected districts should represent a general view of overall Soviet planning and give enough material to formulate universal conclusions. 2. Sociological web survey – to acquire information of residents’ habits and movement in selected territory, by fact. Initially, it helps to choose the most convenient space syntax methods for factual problems and finally, it validates space syntax findings. 3. Space syntax analysis – to figure out causes of spatial residents’ behaviour, thus distinguishing general social-spatial patterns and allowing possibility for prediction. II. ter r Itory S electIon It was decided to conduct the experiment in Vilnius, Lithuania, but the selection of the criteria for exact districts was considered, as the leading goal of this task was to choose rather small areas. Diversity was chosen as a primary objective, but the concept of the term itself changed during the process. The analysis was started by studying the development chronology by decades of the 20th century (1960s, 1970s and 1980s). While this approach initially seemed logical, districts of the same decade appeared to have in common almost only the realization date. The second approach was to group districts by morphotype. This method worked well revealing similarities and individualities of the districts – morphological typicality and uniqueness * Corresponding author. E-mail address: [email protected] © 2020 Milda Sutkaitytė. This is an open access article licensed under the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), in the manner agreed with Sciendo. Architecture and Urban Planning Milda Sutkaitytė, Human Behaviour Simulation Using Space Syntax Methods 2020 / 16 Fig. 1. Selected Soviet micro-districts for the analysis, outline colour by morphotype, fractal analysis on the right [Figure: Author of the Article]. were two further criteria for the selection. Later, (inter-) national acknowledgement (as primary recognition of urban solutions could be reevaluated nowadays) and selected territory continuity (as both inner and inter analysis of districts could be performed) were added to the list. All five criteria supplemented each other, but hierarchy of their significance remained unclear. This is why a simple mathematics-based method – fractal analysis (analyzing raster black and white image by repeatedly dividing its canvas into different-scaled grids of equal-size cells, counting the black dots of containing cells and evaluating the ratio between those result in getting a self-repeatability index) [3] – was taken as the sixth criterion. Three different layers: buildings, streets and buildings, parking lots and streets were separated and analyzed individually. Districts with highest and lowest indexes of each layer were further analyzed. The results gave good representation of the urban development morphology and complexity in each district and, most importantly, possibility to objectively evaluate the overall results. Finally, 4 districts were selected (Fig. 1): • Lazdynai – (inter-) national acknowledgement, fractal significance (low (1.455) in “buildings” and high (1.543) in “streets” aspects), chronology (1960s). • Karoliniškės – morphological uniqueness, fractal significance (high (1.510) in “buildings” and low (1.502) in “streets” aspect), chronology (1970s). • • Justiniškės – morphological typicality, fractal significance (low in all aspects (1.470 – “buildings”, 1.494 – “streets”, 1.676 – “everything”), chronology (1980s). Viršuliškės – territory continuity. III. S ocIologIca l Web S u rv ey Sociological web survey was fundamental for choosing the most relevant space syntax methods at the beginning of the research and validating them at the end. It helped understanding local inhabitants’ movement and making general insights on their preferences to urban spaces. Information from the survey was imported to ArcGIS platform to acquire visually sound maps. The three generated maps are discussed further. A. Map of Favorite Stores The most important map was generated from the survey question: “Where do you usually buy your groceries?” Later, vectors of respondents’ living place and mentioned groceries store were created (Fig. 2). This map obviously exposed that all 4 analyzed districts possess completely different commercial and local movement structure. Justiniškės does not have a well-developed local centre, groceries stores are scattered along the main corridors with strongest concentration in 3 places (Fig. 2: 1, 2, 3). Justiniškės is one of the most remote districts from the city centre, but the newly 85 Architecture and Urban Planning Milda Sutkaitytė, Human Behaviour Simulation Using Space Syntax Methods 2020 / 16 Fig. 2. Map of Favorite Stores [Figure: Author of Fig. 3. Map of Favorite Places [Figure: Author of Fig. 4. Map of Avoided Places [Figure: Author of the Article]. the Article]. the Article]. constructed Vilnius western bypass integrated it into the city core from the outer – west side. It is noticeable, as residents tend to shop in the store outside the district but nearby the bypass (Fig. 2: 4). Viršuliškės have an opposite structure – a very strong local centre (Fig. 2: 5) and a node of secondary importance (Fig. 2: 6), with potential to work as a global centre due to its location besides the main corridor of the whole western Vilnius (Laisvės av.). On the other hand, it does not attract many locals, compared to the primarily discussed store. The reason could be that it is segregated from the main district’s body by two corridors – Laisvės av. and Viršuliškių St. Karoliniškės has a very strong cluster of three stores (Fig. 2: 7), which is popular not only locally but even among the residents of nearby districts. This attractiveness suggests that the indicated hot-spot could become a global centre. Lazdynai represents a district with underdeveloped commercial structure. There are a few nodes (Fig. 2: 9, 10, 11, 12) with some attraction, but the respondents usually use them only for instant purchases while choose to go out of the district for greater shopping. The discussed cluster in Karoliniškės is the most attractive location, but the stores in the city center or anywhere on the commuting way are considered as better options than the local ones. B. Map of Favorite Places The second map was created based on the survey question: “What is your favourite open place within the district?” This question was borrowed from the sociotope mapping method [44] and, like in the previous map, information was uploaded to ArcGIS platform, generating vectors of residents’ living places and their favourite local destinations (Fig. 3). The obtained map revealed that the selected territory by resident’s movement tendency could be easily split into two parts: north (Justiniškės, Viršuliškės) and south (Karoliniškės, Lazdynai). Residents of the south side tend to move out of their districts to the large green bordering territories, while northerners more often stay inside their districts. A couple of main insights could be made from that. Firstly, if physical distance is decent, residents tend to choose larger, more natural environments for their free-time. On the other hand, residents highly appreciate well regenerated inner-district spaces, such as S. Geda alley, Šaulių alley, Šeimų Ppark, East Lazdynai alley (Fig. 3: 1, 2, 3, 4) or spaces having specific self-identity, such as Šimulionio garden (Fig. 3: 5). This confirms that redeveloped spaces are able to conquer residents for pleasing visits and to bring back life to the inner-structure of the city. 86 Architecture and Urban Planning Milda Sutkaitytė, Human Behaviour Simulation Using Space Syntax Methods C. Map of Avoided Places The third map was created according to the survey question about places residents avoid in their districts (Fig. 4). The most interesting correlation is that the places avoided by most users are the same places that were designed as local centres of micro-districts (Fig. 4: red numbers) and remain among the favorite stores even nowadays (Fig. 2: 2, 5, 7, 9). Overall, shopping places and bars are the most alerting places for the inhabitants. The second category of most avoided places appeared to be the forest and parks (Fig. 4: green numbers). This, interestingly, correlates with the Map of Favorite Places and suggests that feeling of unsafety is not overwhelming other qualities that open green spaces possess. The third insight could be that the bigger the street is, the more unsafe it is perceived – the main western Vilnius axis, Laisvės av., with its underground passages, stands out as a particularly obvious example.Analysis of smaller scale places reveals that the yards of apartment buildings are perceived as especially unsafe, and elements such as arches and narrow passages between the buildings play a big role in that. Iv. S pace S y nta x a na lySIS The main concept of the space syntax analysis is that any spatial configuration has inherent relationship with social, economic and cultural aspects and cannot be fully perceived without their understanding [5]. Based on this idea, three different space syntax methods were performed and analyzed individually, as well as validated by comparing information acquired from sociological web survey. D. Segment Analysis: Local Pedestrian Paths Urban structure in urban segment analysis is perceived as a network, consisting of segments of paths or streets and is based on: graph theory, studying simple mathematical structures, containing only two different kinds of elements – nodes and edges, where edges represent relationship between the nodes [5]. To represent local users of networks, metric radius type was used in the analysis as human perception of distances, while moving in well familiar territory, is primarily grounded on physical distances [6]. The analysis of local pedestrian paths network was chosen as a base for humanity infusion to micro-districts, as it can reveal human scale perception of existing spatial configurations. Two different indexes of segment analysis were analyzed: choice and integration. Higher choice values indicate more active places in the network. Higher integration values indicate better reachable places in the network. Three different radius sizes were used, as they could convey diverse usage functionality of the urban spaces – instant, local and global zones, respectively: • 200 metres radius (R200) could be perceived as radius of “here” zone [7], as people tend to not consider other movement options if their destination is within this distance [8]; • 400 metres radius (R400) – pedestrian shed. Based on the New Urbanism principles 5 min walk is a distance people 2020 / 16 choose to walk on foot instead of riding a car [9]. This coincides with the Map of Favorite Stores (Fig. 2) results. • 1000 metres radius (R1000) is maximum comfortable distance for a pedestrian. Based on TOD it is maximum distance of regional transit services or a distance, which pedestrians accept to walk to larger green areas [10], especially if that distance is interesting and useful (e.g., a person can make some shopping on the way) [11]. This coincides with the Map of Favorite Places (Fig. 3) results. Choice Map R200 (Fig. 5). Largest high-value clusters in Justiniškės and Viršuliškės appear nearby pedestrian alleys. A long standing-out stripe appears along Laisvės av. in Justiniškės district. There are later built apartment houses with dense entrances oriented to the street. This hot-spot confirms, that such planning could bring local-users’ activity to a street. Similar clustering situation is in Lazdynai – exposing the remains of pedestrian alleys and territory of highly appreciated modernistic gymnasium. Completely different situation is in Karoliniškės – most of the territory is highlighted and it is hard to point out any separate clusters. This could be because of the overall higher building density in this district. Such structure suggests that Karoliniškės could be well redeveloped for private small spaces, but it could be hard to find a place for joint neighbourhood activities. Choice Map R400 (Fig. 5). Parts of main local transport corridors, some of the street-crossings and major pedestrian alleys inside the neighborhoods start to stand out, while inner-yard structure dissolves to the background. Viršuliškės seems to have concentrated a single core while other three districts have multiple intensive spots. Choice Map R1000 (Fig. 5). Local transport corridors come to the foreground while all inner structure dissolves. Justiniškės and Karoliniškės dominate, while Viršuliškės and Lazdynai seem to have more concentrated and less overall active structure. Residents of Justiniškės and Karoliniškės seem to scatter more when going for longer walks, compared to other two districts. Integration Map R200 (Fig. 6). Dispersity of the values is low and there are no exclusively integrated areas overall. Main living areas of the three northern districts are quite evenly integrated, while Lazdynai lack symmetrical relationship – only three separate zones in southern part show integrity while the north remains completely segregated. Places with highest values of integration and choice parameters of R200 overlay almost identically. Integration Map R400 (Fig. 6). This map looks completely different and exposes clear zones in each district that could be developed as local centres. The map confirms that Soviet micro-districts are very segregated from one another by large streets and green areas – the structure does not have a smooth urban transition. Integration Map R1000 (Fig. 6). Only two intensive areas remain – in Justiniškės and Karoliniškės. These zones overlap with Choice Map R1000 and have a potential evolving to attraction points of the whole western Vilnius. Almost all the territory of Viršuliškės is well integrated, while Lazdynai appears as under-developed. 87 Architecture and Urban Planning Milda Sutkaitytė, Human Behaviour Simulation Using Space Syntax Methods 2020 / 16 Fig. 5. Segment analysis of local pedestrian paths, choice R200, R400, R1000, colour-corrected [Figure: Author of the Article]. Fig. 6. Segment analysis of local pedestrian paths, integration R200, R400, R1000, colour-corrected [Figure: Author of the Article]. 88 Architecture and Urban Planning Milda Sutkaitytė, Human Behaviour Simulation Using Space Syntax Methods 2020 / 16 Fig. 7. Segment analysis of Vilnius city streets, Integration*Choice R7200, colour-corrected [Figure: Author of the Article]. E. Segment Analysis: Global Transit Segment analysis of streets Integration*Choice R7200 (Fig. 7). This Map is generated by mathematically multiplying integration and choice values of the streets’ segments. This analysis draws attention to main logistic corridors and could be used to identify best places for objects of city significance. Part of Pilaitės av., between Laisvės av. and Vilnius western bypass could become such a place as its integration value is one of the highest in the whole city. Another finding is that, probably because of the newly-constructed Vilnius western bypass, former essential western Vilnius axis, Laisvės avenue, has lost its strength in the south. The part, crossing Lazdynai district diagonally, is not chosen often by long-distance users. On the other hand, it helps Lazdynai to regain spatial unity, which due to not fully realized urban project was disrupted. These insights suggest that this part of Laisvės avenue could become a smaller street. In this way, the joint, going through the Park of Fairytales might help overtaking more traffic-load of the transport leaving the city, while maintaining the integrity of Lazdynai. F. Visibility Graph Analysis: Social Control The main idea of the visibility graph is to find direct visual relations between objects based on metric step depth, while taking into account both transit and static usages of the spaces. This analysis supplements segment analysis well, as it includes estimation of shapes and curves of the surrounding urban elements [5]. These inter-visual links between entrances to the buildings are essentially important to ensure social control to the surrounding area, as spaces faced by more doors are perceived safer and are protected from vandalism and other crime acts more, therefore bringing more human activities and vitality there [12]. Social control works as a catalyst to urban liveliness as its existence (presence of habitants) increases the number of users, consequently making the place more public and socially vibrant [13]. This is why balance of open-places privacy levels could stimulate different usage of urban spaces. To improve intelligibility of visually perceived results, metrical range of quantity of social control had to be decided: • 7.5 metres were borrowed from the proxemics field as a margin where private (social) zone finish [14] and interpreted as a maximum distance, which human perceives having control of (Fig. 8: blue colour); • 137 metres are considered as a maximum distance where space is perceived as a place, but not a field yet [15] – “here” but not there yet [7] (Fig. 8: red colour). Everything further should be considered as space with no social control, while everything in-between those distances are semi-private zones (Fig. 8). The acquired map accurately represents typical Soviet urban planning, where main streets lack human interaction and building principles do not create a system of “places”, rather vast empty “spaces” [16]. 89 Architecture and Urban Planning Milda Sutkaitytė, Human Behaviour Simulation Using Space Syntax Methods 2020 / 16 Fig. 8. Visibility graph analysis for social control Fig. 9. Agent-based analysis for pedestrian movement, 400 and 200 metres, colour-corrected [Figure: Author of the Article]. [Figure: Author of the Article]. Large red areas mostly appear in green forested zones or vast fields and do not provide information, which would be unperceived without this map. However, attention should be paid to smaller zones, as they indicate parts of bigger streets or semi-segregated territories lacking social control and need for regeneration. Yellow and green areas often indicate smaller streets, squares, yards, and garage surroundings inside the neighbourhoods and should be considered for most inner-neighbourhood developments. Cyan is the most common colour in the inner structure, while wider, more interconnected blue areas are very rare due to single-side faced entrances, not inter-visible to each other. This map is worth analyzing together with the Map of Avoided Places (Fig. 4) and Map of Favourite Places (Fig. 3)., e.g., Šaulių alley is one of the favorite places (Fig. 3: 2), but is avoided by a lot of people as well (Fig. 4: red 2), as it attracts suspicious persons. Social Control Map reveals that lack of symmetrical relationship could be the reason why. Similar conclusions could be made about most places of avoidance. Creating symmetrical relationship between private and public places by means of smooth transition or transparency could be essential in restoring users’ trust in these places. G. Agent-based Analysis: Movement in Open Space Agent-based analysis is a movement simulation method giving possibility to generate object’s motion in an open-space manner instead of using the existing path network like segment analysis does. The main difference from segment analysis is in the fact that the starting points (in this case – doors to the buildings) can be selected for the agents and their chosen movement in the selected distance or time can be simulated. This is done by implementing simple artificial intelligence – agent – into visibility graph and simulating natural movement by elementary mathematical rules [17]. Agent-based analysis was applied to all districts individually, and the results were combined together into one image only after the analysis in order to acquire a smaller (closer to human) scale grid of 2 metres, therefore the comparison between the districts should be made with caution. The analysis was performed twice for each district with different timesteps in system – 100 and 200, representing 200 and 400 meters respectively, for instant and local movement, as described in segment analysis. Agents were released from entrances to the buildings. 90 Architecture and Urban Planning Milda Sutkaitytė, Human Behaviour Simulation Using Space Syntax Methods • Results of Map of 400 metres are similar to the Integration Map R200 (Fig. 6) of segment analysis, however, clustering of the flow is more noticeable, creating organic spaces of diverse size, configuration and centrality. Space 1 (Fig. 9) could be one example, where segment analysis indicates the axis itself, but agent-based analysis supplements this knowledge by revealing natural spatial configuration of the space that could become a local centre (Fig. 9: 1). Spaces, such as 2 and 3 (Fig. 9) are hardly identifiable in segment analysis at all – the whole area shows great integration, but their centrality is unclear. These are just a few examples that agent-based analysis is capable to recognize ‒ to spot minor spatial differences and to find the most naturally magnetizing centres even in a small scale. Moreover, it reveals a spider-web-like network of how all social hot-spots are inter-connected, that was not visible in segment analysis. Map of 200 metres in most cases suggests similar peak-active places but their configuration is not all the time the same. This information could be useful adapting spaces to different types of users, e.g., Space 4 (Fig. 9) in Map of 400 metres indicates a strong elongated axis, while in Map of 200 metres it highlights smaller intensive space in the north. Furthermore, some areas in Map of 400 metres are over-intensive to extract the main tendency of the zone, while Map of 200 remains more specific (Fig. 9: 5, 6, 7). The congruence of results obtained from both segment and agent-based analyses confirms that the segment analysis is able to convey a pedestrian flow, yet agent-based analysis elaborates a more detailed view. It could be said that the agent-based analysis combines choice and integration characteristics of segment analysis, but supplement them with information from the surrounding obstacles. • • • c oncluSIonS • Sociological web survey proved its suitability to be used as a starting point to review residents’ main behaviour patterns in selected territory and to formulate essential spatial structure problems based on them, as well as to choose the most relevant space syntax methods able to reveal the reasons why such patterns occur. It was important in choosing the right space syntax methods from a huge variety of them, as well as validating their results at the end. Segmental analysis of pedestrian paths revealed places, which are best reachable and most often passed by different kind of users: instant (staying in the yard), local (going for daily needs) and global (going for rarer needs or walk). Identification of such places helps choosing the best functions for them and nearby objects, while lack of closeness can explain their avoidance or abandonment. High overlapping of integration and choice values indicates synergetic logic between pedestrian movement and the existing space configuration in analyzed territory. Standing-out places of both indexes show great potential for becoming a final destination as well as on-the-way stop. Segment analysis of city streets identifies the main city corridors. Parts of streets, which lack closeness and reach, could be considered becoming more pedestrian-friendly. The ones that coincide well with local pedestrian movement, could become transit-oriented development centres. Agent-based analysis represents natural pedestrian flow inside the districts identifying most active places, their size and configuration taking into account the selected obstacles. Nonetheless, the agent-based analysis is able to explain some patterns noted in sociological web survey, such as low elongated clustering in the yards confirming transit oriented urban spaces or urban spaces, which have natural spatial gravitation and are potential for successful development. Visibility graph analysis revealed that the yards and streets of Soviet micro-districts lack social control due to absence of inter-visibility between entrances to the building and balance between private and public spaces. Overall, this kind of map could be used as an easily readable tool for a quick social control comprehension in an analyzed territory. Combination of different space syntax methods provided for the research a particularly complex view. All selected methods were suggesting similar results, but they synergistically supplemented one another – assembling even more detailed picture. Abandonment of any of these methods would negatively influence the overall perception and reduce suggestions for possible problem solutions. r efer enceS 1. • 2020 / 16 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. Zaleckis, K., Tranavičiūtė, B., Grunskis, T., Gražulevičiūtė-Villeniškė, I., Vitkuvienė, J., Sinkienė, J., Armagan Doğan, H. Modernization of Public Spaces in Lithuanian Cities. Berlin: Sciendo, 2021, pp. 71–74, 133–139. https://doi.org/10.1515/9788395793875 Grunskis, T., Šlupinskas, M. Tarybinio laikotarpio viešųjų erdvių transformacijos Lietuvoje. Vilniaus Žirmūnų ir Lazdynų atvejis. Journal of Architecture and Urbanism, Vol. 36, Issue 3, 2012, pp. 209–221. https://doi.org/10.3846/20297955.2012.732492 Chen, Y. Fractal Analytical Approach of Urban Form Based on Spatial Correlation Function. Chaos, Solutions and Fractals, Vol. 49, 2013, pp. 47–60. https://doi.org/10.1016/j.chaos.2013.02.006 Steuteville, R. 25 great ideas of New Urbanism. CNU (The Congress for the New Urbanism), 2018. 204 p. Zaleckis, K. Erdvės sintaksė urbanistinei analizei: koncepcijos, apskaičiavimai ir pavyzdžiai. Vilnius: Vilniaus dailės akademijos leidykla, 2018, 154 p. Hiller, B., Turner, A., Yang, T., Park, H. T. Metric and Topo-Geometric Properties of Urban Street Networks: Some convergences, divergences and new results. The Journal of Space Syntax, Vol. 1, Issue 2, 2010. https://discovery.ucl.ac.uk/id/eprint/18583 Cullen, G. The concise townscape. New York: Van Nostrand Reinhold Company, 1961. 200 p. Penn, A., Turner, A. Space Syntax Based Agent Simulation. Pedestrian and Evacuation Dynamics [M. Schreckenberg, S. D. Sharma eds.]. Berlin, Germany: Springer-Verlag, 2001, pp. 99–114. https://discovery.ucl.ac.uk/ id/eprint/2027 TOD Standard, 3rd edition. New York: ITDP, 2017 [online]. Institute for Transport and Development [cited 01.08.2020]. https://www.itdp. org/2017/06/23/tod-standard/ The Energy Foundation, China Sustainable Cities Program, Calthorpe Associates. Sustainable cities China: Design manual for low carbon development, 2012, 98 p. [online, cited 01.09.2020]. http://www.chinastc.org/ sites/default/files/CSCP_LowCarbonDevelopmentDesignManual_EN.pdf Fang, C., Danlin, Y. Urban Agglomeration: An Evolving Concept of an Emerging Phenomenon. Landscape and Urban Planning, Vol. 162, 2017, pp. 126–136. https://doi.org/10.1016/j.landurbplan.2017.02.014 91 Architecture and Urban Planning Milda Sutkaitytė, Human Behaviour Simulation Using Space Syntax Methods 12. 13. 14. 15. 16. 17. 2020 / 16 van Nes, A., López, M. J. J. Micro scale spatial relationships in urban studies: the relationship between private and public space and its impact on street life. Proceedings of the 6th Space Syntax Symposium (6SSS), Istanbul, Turkiye, June 12–15, 2007. http://resolver.tudelft.nl/uuid:9356f6ae24da-4108-9ef8-51dfca2d2ea2 Whyte, W. H. The Social Life of Small Urban Spaces. New York: Project For Public Spaces, 1980. 53 p. Hall, E. T. The hidden dimension. New York: Doubleday, 1966. 240 p. Blumenfeld, H. The Modern Metropolis: Its Origins, Growth, Characteristics and Planning. Cambridge: MIT Press, 1967. 377 p. Castells, M. An Introduction to the Information Age. The Information Society Reader [F. Webster, R. Blom, E. Karvonen, H. Melin, K. Nordenstreng, E. Puoskari eds.]. London and New York: Routledge, 2004, pp. 138–149. Ståhle, A. Sociotope mapping – exploring public open space and its multiple use values in urban and landscape planning practice. Nordic Journal of Architectural Research, Vol. 19, 2006, pp. 59–71. Milda Sutkaitytė currently is an architect. She received a degree of Bachelor of Architecture in 2014 from Vilnius Academy of Arts and has pursued architectural practice both in private and public sector. Since 2019, she continues Master studies of Architecture in Kaunas University of Technology. Her current research interests include revitalization of post-Soviet urban fabric, urban networking (especially public spaces), and human behaviour prediction based on architectural and urban structures. c ontact data Milda Sutkaitytė E-mail: [email protected] 92