City-scale Digital Twin Prototype for Cambridge

Case Study Overview

The Cambridge City-scale Digital Twin (CDT) project addresses a local policy aspiration of developing a challenge-led Cambridge digital twin strategy, one of the first in the UK. The distinctiveness of the CDT project lies in that its development explicitly adopts a socio-technical perspective, as opposed to a purely technology-driven approach in early smart-city initiatives. As such, the DT Hub case study of the CDT project includes two separate papers, examining the governance and technical perspective, respectively.

What challenges does it solve?

The CDT prototype aims to address the following challenges, a) professional/organizational silos in local policymaking; for example, treating traffic congestion or air pollution as a sectoral issue without considering the interdependence and the coordination with other policy domains; b) unwieldy and often outsourced policy analysis tools (e.g. transport/energy model) hindering the cross-sectoral/departmental collaboration at early stage of policy design; c) limitations of conventional methods for stakeholder engagement and public consultation; d) significant institutional and technical difficulties in data sharing and management; e) the need for cross-disciplinary insight and a joined-up approach for sustainable growth in Cambridge through improving the quality of life and competitiveness of Cambridge.

University of Cambridge has been a key stakeholder in local planning and development. In particular, there has been a productive tradition of using digital models developed by academics for supporting city and infrastructure planning in Cambridge. The CDT project builds on such legacy and aims to explore how the emerging urban data, e.g. from ubiquitous sensing, and a new generation of digital twin models could better support local development initiatives.

How have you integrated data & technology?

The current data used for the Cambridge Digital Twin (CDT) prototype is dispersed between local authorities and external organisations, within which levels of accuracy, spatial-temporal resolution and data standards vary. Therefore, there is a requirement to a) adopt common data standards across the board; b) integrate historical, current and future data and analytical needs into a consistent roadmap that is co-developed with local key stakeholders and citizens; c) identify proper and practical legal procedures for enabling effective data sharing across a wide spectrum of stakeholders and organisations; d) facilitate learning across sectors and scales through incorporating multi-level feedback loops in the design, use and management of CDT tools.

In terms of the selection of technology, it is critical to follow a policy-driven approach, selecting the technology according to specific policy questions and challenges defined in a multi-stakeholder setting. This is in stark contrast with a supply-driven approach where city-level digital twins are framed as a generic technology deployed regardless of local policy context. It is also important to identify the often hidden (at least to CDT developers) constraints of CDT users and owners at the early stage of technical development.

Do you use a user interface to share information?

What digital twins of cities represent and how they are represented requires interdisciplinary insights and participative processes that involve prospective users. The user interface of the current CDT prototype is oriented towards professionals in local authorities and academic users. Extended user interface tailored according to different user backgrounds is to be explored in the next stage. To convey key model outputs to stakeholders, workshops, focus groups and individual meetings have been held by the CDT development team. Workshop participants varied across local government levels (city/district, county, metropolitan, national); the public sector and private sector actors (e.g. University of Cambridge, service providers, consultancies, employers), citizen activist groups (e.g. Smarter Cambridge Transport, CamCycle) and residents (e.g. Federation of Cambridge Residents’ Associations – FeCRA). Such a variety of stakeholder groups is important because it reveals the varying and sometimes conflicting needs for the CDT and the distinct expectations of the user interface.

Our engagement with a wide spectrum of CDT users suggests that there is no single optimum user interface that would meet the needs of all users and at all stages. One critical consideration for designing a suitable user interface is to ensure frequent interaction between CDT model and users, and the frequency of interaction should be informed by the rate of change of the policy examined by the CDT, and further tested with feedback from users.

What outcomes have you delivered?

This is an outcome-based investigation that provides proof of concept and reinstates the need for city level development of digital tools for decision makers, a digital twin model that can cut across several policy areas and allows for quick scenario testing. The current outcomes of both Phase 1 and Phase 2 are outlined as follows.

In Phase 1, to demonstrate possible uses of the CDT model in real-world policy decision-making, two digital transformation scenarios were tested with the CDT model, 1) prevalence of teleworking; and 2) future charging demand for electric vehicles. Both scenarios were chosen jointly with the participants of the 2019 workshop as they serve as good examples of quantifying the cross-sectoral linkages in local infrastructure planning. The teleworking scenario examines two types of impacts of teleworking, a) the potential reduction on commuting flow assuming all teleworkers do not commute on a certain workday; b) the possible mode shift due to teleworking. Key scenario assumption is that the number of teleworking jobs in Cambridge could reach 18,000 by 2051, which accounts for about 13.2% of the total employment in Cambridge. This assumption is developed by considering a detailed breakdown of workplace population by socio-economic background. Note that our assumption (13.2%) seemed radicalat that time, but recent ONS data suggests that in April 2020, 46.6% of people in employment did some work at home due to the pandemic. In terms of model outcomes, new types of workspace (e.g. home/co-working) are blurring the boundary between jobs and workplace – teleworkers search for a suitable place to work and fill that workspace but not necessarily a job.

Planners, developers, architects and employers need to consider new models of providing and managing workspace, not simply reducing the office floorspace per worker, but considering the flexibility and adaptability of space and the changing spatial pattern of commuting demand. The second scenario assumes that all commuting cars to Cambridge as workplace are electric vehicles (EVs). Two variations include 1) EVs charge exclusively at work locations and 2) home-charging only. The results point out that further concentration of employment in Cambridge (See Image 3) would result in soaring charging demand within the city if all commuting EVs would charge at the workplace. Nonetheless the increased electricity demand may be shifted away to other locations in the short term through pro-active pricing policy on charging and/or parking at workplace, as suggested by the comparison between 100% charging at workplace and home place in the figures. To make it possible, it requires better collaboration between transport, housing and energy in local government – the city-level digital twin model may help to bridge such departmental silos.

In Phase 2, the investigation transitioned to a major development site in Cambridge – the Cambridge Biomedical Campus. Using a more focused approach, the outcome of this digital twin was to link multi-sourced data to understand the nuanced car travel demand on site and inform policy design for reducing car dependence. As a complex site with several hospitals, businesses, and research facilities, it is under pressure due to its rapid growth. The outcomes of this phase of CDT development include, 1) a rule-based algorithm for identifying various car user groups on site using both quantitative data and qualitative data; and 2) a comprehensive understanding of the demand, expectations, constraints and concerns of prospect CDT users through extensive stakeholder engagement. Detailed findings on 2) can be in another case study paper focusing on the governance perspective of the Cambridge Digital Twin project. The Phase 2 exercise looked at how intelligence can be drawn from not only digital data but also stakeholder engagement in order to address the need for transparency, participation, and accountability throughout the development process of a city-level digital twin.

A key reference for the CDT project is the Gemini Principles of purpose, trust, and function. From these research investigations, a ‘Future Use’ strategy is being formed around these principles with a key emphasis on the following developments: quality of data, data infrastructure, federation of digital twins, e.g. roads, traffic signals, kerb side, bus/rail networks and legacy systems.

Have you delivered any unexpected benefits?

The development of the Cambridge CDT prototype through a participatory process highlights several issues which future research on digital twins needs to address. First, developing CDTs requires collating data from both conventional (e.g. census) and emerging (e.g. sensory technology) sources. Working with prospective users, however, made it explicit that collecting and linking data from various sources must be driven by specific policy and practical questions. There is urgent need for a shift towards a challenge-led approach for collecting data from, and sharing data with, specific societal actors, due to the resources (financial and human) required and the risks involved (e.g., changing power dynamics, privacy, and security issues).

Adopting to a challenge-led approach and identifying the purpose of CDTs is not straightforward. The participatory process called attention to the difficulties involved in reframing broad policy goals into targets which CDTs can address, and to translating model outputs into relatable and actionable policy insight. A need has, therefore, been identified for prospective users to better understand the functioning and the boundaries of data-driven decision-making support tools (including CDTs) in terms of opportunities, limitations, risks, and uncertainties. Moreover, modelers must engage with a diverse set of societal actors, as well as a wide range of policy alternatives and modelling approaches, to be able to provide contextually relevant and appropriate insights and recommendations.

What lessons have you learn that apply to future work?

The Cambridge Digital Twin project confirms the need for a socio-technical approach for developing digital twins as the next generation of digital tools for supporting city and infrastructure planning and management. From a technical perspective, we follow a policy-oriented, challenge-led approach for designing of the CDT tools. Effective stakeholder engagement since the
onset is critical for defining the purpose and boundary of the CDT tools.

Producing technical outputs is not the end of the CDT development. Instead, urban modellers would have to engage directly with prospective users and incorporate a feedback loop between the model and the users, such that the design and interpretation of the model could be progressively refined and adjusted to the changing policy context and user needs. Given the current development stage of city-level digital twins, we propose two unique roles of digital twins for strategic planning, 1) identifying systemic policy risks and efficiencies at the early stage of policy making, which often caused by a lack of cross-boundary policy coordination); and 2) informing where and when a targeted modelling effort is required, rather than replacing existing, highly sophisticated sectoral models. For city-level digital twins to thrive in policy practice, it is important to re-think the compatibility and complementarity between emerging digital twins and legacy sectoral models.