A Reflection on Medieval Clocks & Digital Twins
What follows is a short reflection on digital twins, medieval clocks, smart city aesthetics and the deviousness of models. The underlying question is: what is the relationship between the device used for simulating the movement of bodies in a system and the system that the device simulates?
The relationship is a metaphor of sorts, where metaphor is understood as something that ‘binds separate things together in a single image’ (Sennett, 1994, p. 79). Together, these two things form a kind of whole, and a reciprocal exchange affects both entities (Black, 1962; Hnilica, 2020). Cities have always generated metaphors that illustrate how they are or should be organized, how they can be cured or how they (should) function. The metaphor then provides an image that is the image of order, of logic, or of form in one way or another.
In this text, I make an attempt to tease out a specific type of metaphor where the ‘bind’ does not directly connect the two things; it does so only in a round-and-about way. This specific kind of metaphor concerns the logic and inner workings of a complex system. The point is that the metaphor is produced through an act of simulation of that system, where someone simulates the movement of bodies within the system and then transposes the logic of the simulation onto the original system. Since the two systems appear to behave in the same way, the simplified logic of the simulation is re-applied on the complex system. One oft-repeated example is that of the Medieval clock. In Medieval timekeeping, the ambition was to create a mechanical device that could capture time, a wheel that would turn one full circle in one day in accordance with the equinox. Already here there is a connection between astronomy and timekeeping. This ambition was articulated by Robertus Anglicus in 1271 (Landes, 1983). The aim of Medieval clockmakers was to better organize the day’s ceremonies in monasteries. In the next stage, in the sixteenth century, elaborate clocks were made that would track a variety of celestial bodies’ movement across the sky. The clock developed into a real-time mechanical simulation of the universe. The real-time simulation helped astronomers and others to believe they had understood the universe; after all, the planets appeared as predicted by the clock. As the clocks were made ever-more intricate, they would, theoretically, become increasingly accurate models of the universe, accounting for and predicting more and more parameters. From here, it was, as the Swedish historian of ideas Sven-Eric Liedman points out, but a short if remarkable conceptual leap to reverse the relation between model and the modelled: if the clock was a functioning model of the universe, would it then not be logical to assume that the universe functioned as a clock? Understanding the universe in this way placed God not as the creator, but as the clockmaker of the universe (Liedman, 1997). Cosmos was reinterpreted according to the logic of the simulation.
I would like to suggest that this type of metaphor, which comes from a simulation of the system itself rather than from an outside object (body, machine) is slightly different. It differs in some respects from the more direct metaphors of the smart city, a topic which has received some attention in recent research (Söderström, Paasche and Klauser, 2014; Hnilica, 2020). There, the two (modern) narratives of the city that is like a machine and/or an organism are explored, along with the notion of the (smart) city that is like an ‘intelligent cloud’. I would suggest that the metaphoric relation I try to capture, what we could call a simulated metaphor, works differently. Rather than two things merging into a single image (as per Sennett), the transposition of the logic of the simulation (clock) onto the original system being simulated (universe) is a peculiar and closed circuit. It is different in that it is so clearly built on the simulation’s resemblance in movement to the system simulated, even though the logic of the system is of a fundamentally different order. It also differs in that it permits us to make predictions about the system’s future movements based on the simulation. This means that we can anticipate future events through the simulation, and transposing the logic of the simulation onto the logic of the complex system places the maker of the simulation in control of the future of the system. We imagine that the logic of the simulation is the logic of the system simulated, which is something we only do if the simulation predicts reasonably well. In this metaphor, there is a dimension of futurity that is based in the simulation. On top of that, there is a dimension of control of the future movements within the system. What is implied is a power in the system rather than a power over the system (the planner-as-doctor who cures the ailing city or the master planner who constructs and tunes the city machine).
This element of control is in real time; it is not separated in a now and then, but is instead continuous, borderless and formless in a way (Deleuze, 1992). This was the logic of Cybernetics, where prognosis and control were central features, and the city was data. In the 1950s and 60s, cybernetically inspired researchers like Jay Forrester attempted modelling ‘urban dynamics’ at the MIT (Forrester, 1999; Light, 2003; Townsend, 2013; Söderström, Paasche and Klauser, 2014). These early experiments were not entirely successful, and they were eventually abandoned.
The smart city movement started with real-time monitoring and control of the cities. Rio de Janeiro was an early poster child where the Mayor and the municipality could monitor garbage collection, traffic, and public transport from a control room. In recent years, the idea of simulating the different systems together in a computer model in what is known as a ‘digital twin’ (coupled with an original twin, which is the material urban realm) has become increasingly fashionable. As in the definitions of the smart city, definitions of what constitutes a digital twin vary, but one working definition is provided by urban planning researcher Michael Batty: ‘a digital twin is a mirror image of a physical process that is articulated alongside the process in question, usually matching exactly the operation of the physical process which takes place in real time.’ (Batty, 2018, p. 817). It needs to be connected to the real-time processes, and arguably it also needs to be able to learn, to better anticipate (Arup, 2019). At the same time, the digital twin is not made to replicate all aspects of the city, as the construction consulting group Arup puts it: ‘digital twins are not necessarily realistic representations, but are rather relevant abstractions of the physical asset.’ (Arup, 2019, p. 16). In this sense, they are intended to remain different from the reality modelled, as Batty points out that if they were (theoretically) to include every aspect they would be inseparable from the system they simulated. Digital twins can basically be of two kinds: reactive (allowing the municipality to react to events, relaying traffic for instance) or predictive (used to plan development and anticipate future events) (Wildfire, 2018).[1] It is the latter type which is perhaps of primary interest in this context. It would presumably be a digital twin only up to a point when it is detached from real time updates; the digital twin cannot be in real time and the future simultaneously (Batty, 2018). Using closed loop data, then, future ‘what if-scenarios’ can be tested: a torrential rain’s effect on the drainage system, how pedestrianizing a street will affect traffic, how a new UDP would affect city systems, and a host of other applications: ‘By building closed-loop city-level data into the virtual city, monitoring, prediction and control of the physical city can be achieved, in order to solve complex problems of the urban lifecycle’ (Arup, 2019, p. 57). It is in this detaching that the model engages with the futurity aspect of the model. As detached, the digital twin becomes a model that behaves in relation to the original twin in the same way that the Medieval clock related to the universe – according to two entirely different logics. Interestingly, it is when the digital twin loses its reality coupling, its real-ness, and becomes a model rather than an approximation that it can be extrapolated and exercise control. The accuracy of these predictions depends on the model’s inclusion of ‘relevant’ data. Hence, the tendency would be to make the digital twin (the simulation) behave increasingly similarly to the original twin (the simulated system).
The gap between the two diminishes, as the digital and original twin become increasingly similar. As a consequence, that which is deemed irrelevant in the construction of the digital twin but which nevertheless may affect events in the original twin, become less and less. In this diminishing gap between digital and original twin, there is the rest: all those things considered unworthy of modelling and/or impossible to simulate or anticipate. Depending on from where we look at the relationship between the digital twin and the original twin, there is a possibility to say either that what this gap contains is the unquantifiable, the lifeworld that remains beyond prediction and ultimately control. Alternatively, of course, looking at the universe as a clock, it is noise. The latter view is, by definition, both abstract and reductive.
The metaphor of the city as a computer dons new garb with the digital twin. The cybernetic dream of mapping and controlling urban dynamics seems to come true. Viewing the city as a computer is no longer a representation but a manifestation of the specific type of simulated metaphor that this text has begun to articulate. The city as a computer arguably becomes the simulated metaphor enacted.
References
Arup (2019) Digital Twin: Towards a Meaningful Framework. London: Arup. Available at: https://www.arup.com/perspectives/publications/research/section/digital-twin-towards-a-meaningful-framework.
Batty, M. (2018) ‘Digital twins’, Environment and Planning B: Urban Analytics and City Science, 45(5), pp. 817–820.
Black, M. (1962) Models and metaphors: studies in language and philosophy. Ithaca, N.Y.: Cornell Univ. Press.
Deleuze, G. (1992) ‘Postscript on the Societies of Control’, October, 59, pp. 3–7.
Forrester, J. (1999) Urban Dynamics. Waltham: Pegasus Communications.
Hnilica, S. (2020) ‘The Metaphor of the City’, in Figueiredo, S. M., Krishnamurthy, S., and Schroeder, T. (eds) Architecture and the Smart City. Abingdon: Routledge, an imprint of the Taylor & Francis Group, pp. 68–83.
Landes, D. S. (1983) Revolution in time: clocks and the making of the modern world. Cambridge: Belknap P. of Harvard U.P.
Liedman, S.-E. (1997) I skuggan av framtiden: modernitetens idéhistoria. Stockholm: Bonnier.
Light, J. S. (2003) From warfare to welfare: defense intellectuals and urban problems in Cold War America. Baltimore: Johns Hopkins University Press.
Ng, A. (2020) 7D Vision, E-flux. Available at: https://www.e-flux.com/architecture/software/337517/7d-vision/ (Accessed: 28 April 2021).
Sennett, R. (1994) Flesh and stone: the body and the city in Western civilization. London: Faber.
Söderström, O., Paasche, T. and Klauser, F. (2014) ‘Smart cities as corporate storytelling’, City, 18(3), pp. 307–320.
Townsend, A. M. (2013) Smart cities : big data, civic hackers, and the quest for a new utopia. New York: W. W. Norton & Co.
Wildfire, C. (2018) How can we spearhead city-scale digital twins? Infrastructure Intelligence. Available at: http://www.infrastructure-intelligence.com/article/may-2018/how-can-we-spearhead-city-scale-digital-twins (Accessed: 19 August 2021).
[1] On a side note: a similar development is taking place on the level of the individual building through BIM, building information modelling being extended to include building management (Ng, 2020).