World Cities Summit 2022 Science of Cities Symposium - Panel 1: Complexity Science for Adaptive & Sustainable Cities

From top-left: Dr. Sam Joyce (Assistant Professor, Singapore University of Technology and Design), keynote speakers Professor Luis Bettencourt (Director, Mansueto Institute for Urban Innovation of University of Chicago) and Mr. Sanjeev Sanyal (Secretary, Government of India), Mr. Nishant Kumar (PhD student, Singapore-ETH Centre), Ms Anjanaa Srikanth (PhD student, Singapore University of Technology and Design), Dr. Lock Yue Chew (Associate Professor, Nanyang Technological University), and Professor Thomas Schroepfer (Professor, Singapore University of Technology and Design) presented their research and shared their global policy insights in the first panel of the inaugural Science of Cities Symposium on Complexity Sciences for Adaptable and Sustainable Cities. Keynote speaker Associate Professor Ying Long from Tsinghua University and Beijing City Lab was also present virtually.

Panel 1: Complexity Science for Adaptive & Sustainable Cities

1. The future of cities is about how cities see and trust themselves as a complex system that evolves to meet its people’s needs

A city is a complex system comprising many moving parts that continuously interact with each other in unpredictable ways and through feedback loops, leading to unintended consequences. This contrasts with a complicated system such as car engine systems where the component parts interact in pre-determined ways and do not evolve presumably.

Influenced by Le Corbusier’s planning ideology that were based on Newtonian mechanics, cities in the 20th Century were regarded as “machines for living” and designed in overly comprehensive and rigid manners. In his keynote address, Mr Sanjeev Sanyal put forth that instead of planning cities with mechanical top-down views, cities would better evolve with systems of feedback loops that are enabled and informed by the effective use of big data. He illustrated with a comparison between Chandigarh (India) and Singapore – in 1965, the former was a fully-fledged city while the latter was a newly-independent city. However, over the past five decades, Singapore had progressively moved up the value chain from a labour-intensive manufacturing city in the 1960s to venture into the electronics industry in the 1980s, and subsequently ventured into financial markets, entertainment, education and high-end industries from 1990s onwards. In contrast, Chandigarh remained as an administrative city throughout the same period. While some cities in India managed to evolve, the city infrastructure struggled to keep up with the evolution as there lacked a proper scientific approach to facilitate targeted interventions to let the city evolve according to the needs of the time.

Professor Luis Bettencourt added that cities will need scientific frameworks to solve urban problems such as energy transition and environmental or equity issues that cut across disciplines. A theoretical framework that can apply empirically throughout history and extrapolate into the future is illustrated in Figure 1 below. We need to acknowledge cities as a system of systems (e.g., political, economic, energy, transportation systems, among others) that are made up of interdependent networks, heterogeneous needs, costs, benefits and trade-offs, and information which is critical for cities to evolve over time. It is also important for cities to have a social contract for people to come together for problem-solving across scales (individual, household, neighbourhood, city levels).

Figure 1: A theoretical framework to guide the thinking of future sustainable cities (Source: Luis Bettencourt)

“Rather than have mechanical top-down views on how cities would evolve, it would be much better to have systems of feedback loops where we use data such as satellites, to intervene and manage the process.”

Mr Sanjeev Sanyal
Member, Economic Advisory Council to Prime Minister of India; Secretary, Government of India

“The future of cities has a lot to do with how it sees itself as a complex system, how it trusts its systems and how it has a framework to guide itself to imagine its future more scientifically.”

Professor Luis Bettencourt
Professor & Inaugural Director, Mansueto Institute for Urban Innovation, University of Chicago

2. Three pathways to better harness technologies to promote urban development: City Laboratory, New City and Future City

In the Fourth Industrial Revolution characterised by the emergence of disruptive technologies and the ubiquity of big data, sensor networks, Internet of Things, etc., Associate Professor Long Ying shared three pathways to understand the relationship between technologies and cities (Figure 2).

Firstly, technological progress is promoting the establishment and development of urban science, through providing new data and methods that can help planners and researchers understand city patterns. In a “city laboratory”, the time is ripe for us to use active sensing approach (e.g., mobile sensing, stationary sensing, collaborative sensing networks) to collect new data to uncover city patterns and evaluate behavioural changes.

Secondly, technology promotes social and urban changes as it provides opportunities for a “new city”, where there are new ways of experiencing the urban spaces, such as (i) facilitating mixed function spaces where online and offline boundaries are increasingly dissolved, (ii) platform-based urban governance that is enabled by the digital operation of spaces, and (iii) integration of virtuality and reality where digital innovation augments the design of urban spaces.

Thirdly, the pathway of future city design is at the nexus of physical space, information space and social space. Hence, digital innovation must be integrated with and considered as part of efforts for spatial intervention and place-making, when developing smarter future cities.

Figure 2: The concepts of digital innovation, spatial intervention, and place-making for future city design. (Source: Long Ying)

3. High-density liveable cities require new integrated planning and design strategies that acknowledge their complexity and emergent patterns

In land-scarce and high-density cities like Singapore, public and common spaces that were traditionally located at the ground level are increasingly sited on elevated levels in the form of rooftop gardens, sky terraces, etc., and integrated with residential, retail and other commercial developments. The impact of these vertical spaces on the patterns of human movement and their space usage at the building scale and beyond had to be understood, to provide a more scientific basis for planning and design more vertically integrated developments for higher population densities while maintaining liveability.

A complexity science-based approach that harnessed spatial network analysis and an active sensing approach could be adopted. Using Kampung Admiralty (a vertically integrated public development in Singapore), Professor Thomas Schroepfer illustrated a systematic post-occupancy socio-spatial network analysis framework that combined (i) qualitative architectural analysis, (ii) 3D spatial network analysis that mapped buildings and its amenities as nodes and superimposed with socio-spatial information derived from human movement sensor data, and (iii) correlation analysis of actual performance with spatial network patterns. While providing early insights on the design intent of Kampung Admiralty, the methodology also served as a basis for larger urban studies that seek to identify emergent patterns of human mobility and temporal co-presence networks in high-density built environment.

Evolutionary park system is an example where socio-spatial network analysis combined with sensor-based analysis could be further applied to a larger scale. Considering that parks are agents of urban vitality, Ms Anjanaa Devi scaled up the above-mentioned methodology to address the knowledge gap of scientific assessments of spatial accessibility and pedestrian connectivity on a temporal basis for one-north park in Singapore. Specifically, spatial network analysis was applied to predict the effect of one-north park’s topography on pedestrian movement patterns based on a 3-D pedestrian model of the subzone. This was complemented by empirical sensor-based studies of pedestrian movement, which was further co-related with the spatial network analysis to determine the level of integration of the park with its larger

Beyond understanding the locational preferences of humans in space, a complexity science approach can also be applied to understand the spatial patterns of firms’ locations across sectors and the spatial linkages between businesses and industrial activities. In his presentation, Associate Professor Chew Lock Yue introduced four key methods (i) agglomeration (to distinguish different clustering patterns and measure the extent which firms of the same category would cluster or co-locate, (ii) region index (to distinguish sectors that are concentrated versus sectors that are more distributed relative to a region of interest (e.g. CBD area); this can help reveal whether a sector is sticky or diffuse; a more diffused sector could better support a polycentric development, (iii) co-occurrence and co-change (to understand the spatial relationships/inter-dependencies between industries and quantify the tendency of industries making similar changes in location choice over time), and (iv) Central Area (CA) Index (to measure the affinity of a firm of a particular type to situate in the central area relative to the baseline of all firms).

4. As cities are unpredictable, it is important to use real-time data to plan for optionality rather than to know what exactly would happen next

Singapore and many cities in the world are today in a much better position to manage the complexities of our built environment due to the changing possibilities from data. Where earlier quantitative studies were methodologically challenged by the lack of data, the world today is one of data abundance and increasing computational capacity.

As Mr Sanjeev Sanyal put it, “The most important thing to think about is the speed at which change happens and be able to use all kinds of new data to encourage or discourage the changes that city leaders hope to see.” To that end, it is essential for cities to invest in real-time data and data collected from mobile phones, social media, etc. to sense and respond to the immediate happenings and anticipate future development trajectories.

For example, Mr Nishant Kumar (PhD Researcher, Singapore-ETH Centre), utilised real incident and congestion data collected from road networks in Singapore to understand the relationship between the topology and resilience of road networks, and further used congestion duration as a proxy dataset to evaluate the network resilience of roads. With the use of real-time data, his research shed some light on the methodology to quantify road network resilience through a grid-based segmentation strategy, and if a robust segmentation of the urban space could help cities better segregate the low and high resilience segments of the road network, so that interventions could be more targeted.

Furthermore, city planners often have to make plans and policies under conditions of incomplete information and uncertainty. The challenge for the planner is to make plans that are robust while having constant transition management, even as assumptions, conditions, and technologies change. For example, in the context of airport planning, as there is a long gestation time for airports to be realised, cities would have to consider in the present-day its future land-use demands and projected aviation trends in relation to its GDP and population size. For example, Dr Sam Conrad Joyce’s research used a combination of network theory, Central Place Theory and PageRank analysis, to explore how the symbiotic relationship between global aviation networks and urban centres might impact the projected changes in population and GDP, and thus drive new aviation demand profiles, location links, and even new airports.

Writer’s Bio:

Tan Yi Xuan
Manager (Research)
Centre for Liveable Cities

Yi Xuan is a researcher at the Centre for Liveable Cities, focusing on Complexity Science for Urban Solutions, Future Towns and Districts, and Smart Cities.