Bringing the Social City to the Smart City

Lead Research Organisation: University of Leeds
Department Name: Sch of Geography

Abstract

Technological developments, such as the rise in GPS enabled devices and Web 2.0 technologies have created social transformations in how we connect and share information through the mass uptake of smart phones and social media platforms (Croitoru et al, 2014). This new generation of mobile technologies work as individual sensors capturing data on a wide range of human behaviours that have been previously hidden. These include data on individual movement, preferences and opinions. Understanding these behaviours is crucial if we are to create a joined up approach to simulating how cities breath and grow. However, considerable work is required in adapting and developing new technologies from machine learning to extract behaviours which can be embedded into cutting-edge modelling techniques. Creating this bridge between 'big' data representing the 'real' world, and simulations producing alternative versions of reality is of value to both academics and policymakers looking to develop new solutions to many of the challenges that today's cities face. To do this we need to understand how factors within the "Social City" (the impact of individual movements and decisions) play out every day in the "Smart City" (data collected from fixed sensors on for example, traffic counts, air pollution or movements of populations).

However, standard "Smart" City understanding assumes previous flows (e.g. traffic at a specific time of the week, energy requirements or pollution levels over a 24-hour period) will be replicated into the future, lacking both adaptability (how does this alter if a major event in the city is happening?) and predictive power (what is the impact on health if all petrol and diesel vehicles are banned?). This disconnection between the Smart and Social city means policymakers are unable to obtain answers to complex interrelated questions such as: what is the optimal transport infrastructure to promote healthy behaviours and reduce the City's carbon footprint?

Being able to answer these questions is increasingly important as cities are facing significant challenges associated with the pressures from rapidly increasing urban populations. These include improving water and transportation infrastructure, air pollution and waste management as well as provision of adequate housing, energy, health care, education and employment. These pressures on future cities has brought the Smart City agenda to dominate many government initiatives with governments and policymakers looking to new forms of (big) micro data to provide innovative solutions to these challenges. While many models have been developed to forecast future transport, housing or healthcare initiatives, most uses are purely empirical: they lack any consideration of the social processes behind the individual generating the data or the impact of their actions and decisions. This Fellowship will explore how machine learning inspired tools can be used to recognise such emergent patterns and processes within micro-level data sources such as data on how individuals move and use city spaces. Along with the additional methodology of Agent Based Models, which allows smart and social city data to be readily combined, this suite of methods will be explored through looking at the case-studies of (i) the impact of air pollution on individuals and (ii) urban mobility. This work will fundamentally transform our ability to show how the social elements of the Smart City can be recognised and understood, and how to bring 'lived experience' to the analysis of Smart City data.

Planned Impact

One of the key elements of the Fellowship is to develop economic and societal impact through engagement with a wide range of stakeholders. The Fellow will be based within the Consumer Data Research Centre (CDRC). The CDRC has established a network of partnerships that the Fellow will thread together over a number of activities. There are a wide range of active partnerships within CDRC including Callcredit, Axciom, Sainsbury's, Zoopla, Whenfresh, Heart Research UK, M&S, Virgin Media, YouGov and Link. Access to data and stakeholders from these and other CDRC partners will be available to the Fellow. Engagement with external partners is facilitated through events including an annual User Forum, which brings together more than 60 commercial partners of CDRC.

How will they benefit?
Research project (RP) 1 and PhD 1: Creating tools to uncover hidden emergent trends and processes in spatial data. Can new tools give greater insight to explain why a phenomenon is prevalent in one area, but not another? Can new data tools reveal new insights into current data, and uncover gaps? Beneficiaries from this work include organisations that need to probe deeper into spatial data to understand the interplay of emergent patterns over time. Examples: Local government planning and transportation departments, public health organisations.
RP 2 and PhD 2: Testing methodologies through ABM: ABM gives the opportunity to bring together Smart and Social city data through simulating individuals. This aspect of the work will create new understanding about the utility of new social media data sets and how they can be readily combined with Smart data. This will create impact for those beneficiaries who are interested in both simulating and understanding individual behaviour and the consequences of individual decisions. Examples: Retail (commercial) and health organisations, local and national government.
RP3: Application to case-studies. This part of the Fellowship has the greatest potential impact as the work can be readily applied to specific case-studies to answer questions about the consequences of new policies or interventions. Examples: Can the transportation infrastructure be optimised to encourage active travel and city spaces to be more readily used? Whilst methodological innovations will have the greatest short-term impact on researchers in academic institutions, the medium to long-term impact of this work will feed through into research orientated public sector organisations, such as those interested in health behaviours, mobility patterns, impact of new infrastructure and new economic initiatives. Examples: Town and Country Planning Association; West Yorkshire Combined Authority.

How will I ensure stakeholders will benefit from the research?
Stakeholder advisory meetings: Creation of a stakeholder group, the purpose of which is to promote interactions and exchanges. Meetings will be scheduled for the beginning of the Fellowship and each subsequent year to allow time to shape research questions and identify additional case-studies.
Creation of new resources including training courses: Open access code, online training materials will be made available to professionals and academics alike. Workshops will be held to train stakeholders, academics and CDT PhD students in new methods and data, thus impacting on research capacity and skills.
Seminars: Turing and CDRC both offer seminar programmes at which research findings from this Fellowship will be disseminated. Webinars will also be used to increase the impact and range of engagement with different stakeholders.
Further funding applications: Engaging with stakeholders and building bridges is of prime importance for future funding opportunities. At month 18, a research incubator event will be held at CDRC. This will involve Turing Fellows as well as invited stakeholders and researchers

Publications

10 25 50
 
Description Work has been focused in three areas:
(i) developing methods for understanding movements in space-time data (where and how do people move around cities). We have developed network methods (published) and work is currently undergoing assessing visualisation and ML methods.
(ii) initial work has been published looking at understanding causality from the perspective of statistical epidemiology - why and how does something change in a population. This is novel work that links together individual-based modelling with causal inference methods. This work will continue and tools will be produced.
(iii) developing methods for quantifying uncertainty. Research has been published on methods for understanding uncertainty in agent-based models and a grant submitted on this topic. Ongoing work will look at novel methods to improve uncertainty in individual-based models (particle and Kalman Filters) as well as emulators.
Exploitation Route All the code etc that will be developed will be placed on Github and I am working with the Turing RSE group to create reusable code wherever possible. I am actively collaborating and disseminating the work through multiple channels.
Sectors Aerospace, Defence and Marine,Digital/Communication/Information Technologies (including Software),Energy,Environment,Financial Services, and Management Consultancy,Healthcare,Retail,Transport

 
Description Quantifying Utility and Privacy Preservation in Synthetic Populations (QUiPP)
Amount £408,611 (GBP)
Funding ID TPS2019\100019 
Organisation Alan Turing Institute 
Sector Academic/University
Country United Kingdom
Start 10/2019 
End 03/2021
 
Description Real-Time Advanced Data assimilation for Digital Simulation of Numerical Twins on HPC (RADDISH)
Amount £400,000 (GBP)
Organisation Alan Turing Institute 
Sector Academic/University
Country United Kingdom
Start 10/2019 
End 03/2021
 
Title Network Analysis Visualisation 
Description The tool that is being developed is in two stages: network science tools adapted to handle large databases and (ii) adaption of visualisation tools for understanding what's happening inside of the data. My plan is to pass these tools to the Turing for a Data Scientist to create reusable code. 
Type Of Material Improvements to research infrastructure 
Year Produced 2019 
Provided To Others? No  
Impact The tool has not yet been completed. 
 
Title Particle filters for agent-based models (uncertainty handling) - RADDISH 
Description Particle filters are being explored as a method to emulate large numbers of agents and to quantify uncertainty in this approach. 
Type Of Material Improvements to research infrastructure 
Year Produced 2020 
Provided To Others? Yes  
Impact No notable impact as of yet, but this research has been linked to the Alan Turing TPS RADDISH project. This involves creating a real-time warning system for hazards such as tsunamis. 
 
Title Quantifying Utility and Privacy Preservation in Synthetic Populations (QUiPP) 
Description This tool is aimed at creating individual-level synthetic populations from sensitive data sets that preserves the internal structures and preserves individual privacy. 
Type Of Material Improvements to research infrastructure 
Year Produced 2020 
Provided To Others? Yes  
Impact This tool is still in its initial phase. We are hoping to create a demonstrator in the next few months then work with health and financial data providers. 
URL https://github.com/alan-turing-institute/QUIPP-pipeline/projects
 
Title A dataset of dockless bike sharing trips and road network for Nanchang, China 
Description The dataset includes the data and scripts in the paper " Yang Y, Heppenstall A, Turner AGD and Comber A. A spatiotemporal and graph-based analysis of dockless bike sharing patterns to understand urban flows over the last mile. Paper accepted for publication in Computers Environment and Urban Systems (July 2019)". It contains (1) Road Network and dockless bike trips of 10 weekdays in Nanchang, China (2) Scripts used to produce the result in the paper. 
Type Of Material Database/Collection of data 
Year Produced 2019 
Provided To Others? Yes  
 
Description Modelling Uncertainty 
Organisation Improbable
Country United Kingdom 
Sector Private 
PI Contribution This work looks at quantifying uncertainty through the use of probabilistic programming for improving the forecasting accuracy of simulation tools. This directly feeds into the work undertaken by the Fellowship and links to other funded work.
Collaborator Contribution They have funded two intern projects and a PhD studentship as well as giving full access to their probabilistic framework (Keanu).
Impact So far this work has resulted in the developed of two conference papers and software development.
Start Year 2018
 
Description Attendance at the monthly Turing Urban Analytics group 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Other audiences
Results and Impact Monthly meeting and exchange of ideas about research funding and development of ideas for the Turing Urban Analytics programme.
Year(s) Of Engagement Activity 2019,2020
 
Description Turing workshop on reproducibility 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact Co-organiser of the "Reliability and reproducibility in computational science: Implementing verification, validation and uncertainty quantification in silico" workshop.

As a result of this event, I discussed methods with Prof Peter Challenor (Exeter) and wrote and submitted a Turing grant on Uncertainty.
Year(s) Of Engagement Activity 2020
URL https://www.turing.ac.uk/events/reliability-and-reproducibility-computational-science
 
Description Workshop on modelling crime demand 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact Workshop to engage policymakers (police) to understand their requirements for a police demand tool. I am involved in this as a Co-investigator, solicited for my expertise in ABM and demand modelling.
Year(s) Of Engagement Activity 2020