Urban experimentation and smart cities: a Foucauldian and autonomist approach

ABSTRACT This paper provides an analytical framework capable of critically analysing the currently hegemonic phenomenon of smart cities in the contemporary metropolis, particularly concerning their assumptions and potential effects. Using a critical perspective through the scope of governmentality studies and autonomist literature, an analytical framework is presented that allows for the comprehension and critical analysis of the phenomenon in question. Both this theoretical review and the analytical framework will form the basis for a critical analysis of a case study in Dublin, Ireland, taken as an example of the diversity of phenomena and issues to be addressed in this paper. The case study concerns a project by See.Sense, a company located in Dublin, which involved the introduction of 500 bike lights with Internet of Things (IoT) and digital technology in its urban space, this being an example of urban experimentation within smart cities. The paper concludes by defending that the phenomenon in question constitutes not only a form of technological mediation and economic production but also a transformation of urban space and its subjects and population, implying a structuring of social practices and relationships.


INTRODUCTION
The aim of this paper is to present a critical and comprehensive analysis of the currently hegemonic phenomenon related to smart cities, especially the forms of digital ubiquity and mediation in the contemporary metropolis. This theoretical discussion aims to articulate studies on Governmentality with autonomist literature, considering both are linked and make a valid and critical contribution to the objectives of this paper. Regarding the concept of Governmentality, the intention is to explore its primary meaning and implications, much like Michel Foucault did (Foucault, 1988(Foucault, , 2007. However, we also explore this concept while considering other theoretical contributions concerning more recent phenomena, as is the case of the concept of algorithmic governmentality. By using autonomist literature, we can critically explore questions concerning the concept of Governmentality, and to intersect them with a political, economic, technological and territorial perspective. It is considered that the rationale produced by this perspective on urban space and metropolis is useful for a critical analysis of the transformation processes of urban and metropolitan spaces. Also, when combined with the reasoning highlighted in governmentality studies, it allows for the construction of an analytical framework considered useful for the comprehension and critical analysis of the currently hegemonic proposals and expressions regarding smart cities.
This discussion serves as the basis for the analysis of a case study concerning the use of the See.Sense ICON product in Dublin. This analysis allows us to exemplify and explore the ideas and hypotheses presented above. The project emerged from a collaboration between the See.Sense company, dedicated to technology and urban mobility, and Dublin City Council. As a result of this collaboration, 500 cyclists from Dublin participated in a project for digitally mediated mobility in Dublin, with gains for both See.Sense and the public administration. The analysis of this project allows for a critical discussion of the relationship between forms of governmentality, urban experimentation, and technological and digital mediation in the contemporary metropolis. The reference to the See.Sense case aims to explore the discussions presented here, rather than to describe the process itself, allowing for an exploration of the dynamics around the phenomenon of urban experimentation associated with smart cities.
The paper is structured as follows. The next section addresses the concept of governmentality, exploring its various dimensions. The third section introduces a set of post-Marxist and autonomist ideas regarding economics, politics and technology, forming a summarized vision of the perspectives in question and exploring their relationship with previous literature and with the objectives of this paper. The fourth section is designed to articulate the theoretical perspectives presented here with autonomist discussions about the metropolis and the phenomenon of smart cities, concluding with an analytical framework for studying smart cities. The fifth section presents the case study, offering both a contextualization of the urban policy of the city in terms of smart cities and an exploration of this project through the lens of the main categories identified in the analytical framework.
The analysis of this case is based on the relationship between the theory presented and the empirical findings. On 22 June 2017, a semi-structured interview was held in Dublin with a See.-Sense employee. The interview focused on the key aspects of the See.Sense project in the city. In addition, a qualitative content analysis was performed on the See.Sense website 1 as well as on the Smart Dublin website. 2 The purpose here was to identify and analyse meaningful information on these websites from the categories worked on in the analytical framework, and to take into consideration the context and object of research, following the indications present by Kohlbacher (2006) about qualitative content analysis methods. This analysis was carried out at different stages, starting in May 2017 and ending in November 2019, to check possible website and project updates. The information available on the websites was triangulated with other information available in some other Irish newspapers/websites, as well as other institutions related to Smart Dublin and See.Sense. With this triangulation, it was possible to organize and structure information considered to be more consistent and meaningful to the objectives of this paper. By using such methods, it was possible to collect information about the See.Sense ICON product, the experimentation and collaboration project carried out in Dublin by See.Sense, as well as the intentions and gains of See.Sense and of the local authorities of the city.

GOVERNMENTALITY
The concept of Governmentality, introduced by Foucault and later developed in governmentality studies, has mainly been addressed with reference to historical processes, analysis of power technologies that affect individuals and/or populations, or in critical analyses of institutions and forms of government. Given the objectives stated for this paper, the intention is to take a different approach and present another application of this concept.
Foucault's initial theoretical questions concerned the processes and arts of government from a historical perspective, referring to a process of 'governmentalization of the state' which marks the history of Western modernity (Foucault, 2007). Even highlighting the various intersections and overlapping the different paradigms and historical periods in which they become hegemonic, Foucault approaches this process by setting out from its effects on the territory (paradigm of sovereignty), on the body of individuals (paradigm of discipline) and on the population (paradigm of security). 3 This process of governmentalization of the state highlights the way in which, historically, the 'introduction of the economy' occurred in the sphere of 'political practice' and in the 'reason of state'. According to Foucault, this introduction led to a change in the very meaning of the concept of economics, 4 establishing a 'field of intervention'. This process is associated with a conception of government as 'the right disposition of things' for certain ends (government, formation of subjects and populations, among other possible ends) (Foucault, 2007). Nikolas Rose's interpretation of the concept of governmentality is useful for the objectives of this paper, stating that governmentality refers to "the ways in which those who would exercise rule have posed to themselves the question of the reasons, justifications, means and ends of rule, and the problems, goals or ambitions that should animate it" (Rose, 1996, p. 41-42). Governmentality is considered to be a mechanism used to represent, interpret and intervene in reality in such a way that it becomes 'accessible/docile' with a view to a given end and/or political/governmental program (Joyce, 2003, p. 3, as cited in Rose, 1996. However, considering the aim of the research proposed here, it would be necessary to establish not only the use of the economy and diverse apparatus as a means of government, but also the use of techniques and technology to this end. Essentially, it is questionable whether interventions in the field of smart cities and the forms of digital ubiquity and mediation arising in the contemporary metropolis do not also represent ways of conceiving and interveningthat is, epistemes and practicesin what is considered to be reality (Joyce, 2003). Thus, it is important to question the ways of representing and intervening in urban space, population, infrastructures and physical environment. And it is also important to contemplate possible rationales of instrumentality considering the objectives of economic competitiveness, technological innovation and/or ways of controlling and producing specific subjectivities. Taking this into account, the questioning presented in this paper relates to the possibility of economics not only constituting a form of exercising authority, but also of it extending to the forms of digital mediation that have intensified with the most recent transformations and technological innovations.
Moreover, the way the concept of governmentality was subsequently developed by Foucault receives a new and a useful meaning within this paper. Focusing its analysis on the processes of subjectivation, the topic of 'technologies of the self' (Foucault, 1988) allows a different approach to be taken regarding the relationship between dichotomies such as those of agency and structure, individual and collective. Foucault identifies four major types of technologies used by human beings for knowing and taking care of themselves: technologies of production, technologies of sign systems, technologies of power and technologies of the self. Highlighted among these technologies is the relationship formed between the so-called technologies of power and technologies the selfa relationship that, for the author, constitutes a form of governmentality. These are two processes that could be co-constitutive, and lead to the possible relationship between technologies of power created with the intention of structuring individual conduct towards a given end, and technologies of the self that can involve processes of subjectivation that may or may not fit with the assumptions and intentions of technologies of power. That is, it is possible to consider the way a given individual conduct can adjust to and/or reproduce a given hegemonic conception or a given inequality in terms of power relations. And, at the same time, following an immanent view of power relations, explore the possibilities of resistance that can be found even in hegemonic and asymmetric situations and relations.
Essentially, this is the continuation of the relational, positive and productive conception that Foucault had developed about the concept of power. For the author, power is not viewed as a thing, but rather as a process, a network where power circulates between things and in which 'power is exercised only when it is put into action' (Foucault, 2005, pp. 21, 34-35). For Foucault, in-force relations subjects are not the merely passive object of power, since power circulates between the subjects themselves because of their actions (pp. 21, 34-35). Power is understood as relational and productive, acting on possibilities and structuring actions and subjectivities: '[i]t is a total structure of actions brought to bear upon possible actions' (Foucault, 1982, p. 789). Considering power relations as a potentially productive means, there is a necessity to consider its relational dimension, given that: a power relationship can only be articulated on the basis of two elements which are each indispensable if it is really to be a power relationship: that 'the other' (the one over whom power is exercised) be thoroughly recognized and maintained to the very end as a person who acts; and that, faced with a relationship of power, a whole field of responses, reactions, results, and possible inventions may open up. (p. 789) This is an assumption that should be considered in the analysis of power relations, and which we will seek to take into account throughout both the construction of the analytical framework and the case study analysis.

GOVERNMENTALITY, ECONOMY AND TECHNOLOGY
Post-Marxism and governmentality: a short introduction The relationship between Governmentality and Economy, identified in the historical perspective presented by Foucault, is also present in various critical and post-Marxist economic perspectives.
An example that allows for a summary introduction to this line of research is the work of Pierre Macherey. By invoking the Foucauldian concept of biopower, the author highlights the way capitalism would not be possible without the 'controlled insertion of bodies into the machinery of production and the adjustment of the phenomena of population to economic processes' (Macherey, 2015). The author's argument is that the development of capitalism would not be possible without a form of governmentality based on a conduct of bodies, subjectivities and populations. A process that implies the continued use of technologies of power for the 'optimization of strengths, attitudes and life in general' (Macherey, 2015), at the same time seeking to ensure a continued accumulation and adjustment of the bodies to capitalism that implies a reproduction of unequal power relations. Therefore, it becomes possible to consider economics as a means to government ends, that is, as a form of governmentality. Macherey highlights the existence of an 'economy of forces' which is closely linked to an 'economy of persons', implying the structuring and control of the body of the subjects for the reproduction of such economic system (Macherey, 2015).

GENERAL INTELLECT AND COGNITIVE CAPITALISM
The reflections of Hardt and Negri (2001, 2004, 2009) and Matteo Pasquinelli (2013, 2015b allow us to establish a connection between autonomist thinking and its reinterpretation of Marx's thought, Foucault's work on governmentality, and a critical reflection on the role of technology in contemporary capitalism. To summarize, the key idea here is based on the disruptive potential of technological transformations and the emergence of what is known as 'general intellect'. It is argued that there is a tendency in which the whole of society, including the diverse spheres of life, would be taken as the main means of innovation and production of economic value. Something intensified with technological innovations and which, following a Foucauldian perspective, implies a structuring of subjectivities, relationships and social practices considering the possible ends to be attained. This question is analysed in depth by Hardt and Negri in their trilogy Empire-Multitude-Commonwealth (2001,2004,2009). According to them, what they call bio-political production (which includes the various dimensions of life and the above-mentioned 'general intellect') is spread throughout society, particularly in the contemporary metropolis. Essentially, it is the idea that economic production, by including aspects and dimensions traditionally considered to be outside of that sphere, is no longer confined to certain places (such as the factory or the office), time (of work) or relations (traditional work ones, marked by the existence of a salary and a legally certified contract). In addition, such production is essentially collaborative and carried out collectively, in common, while being dependent on affective, relational, immaterial and linguistic dimensions that cannot be altogether individualized. These dimensions move throughout society and the metropolis and are captured for the (re)production of value and capitalism in general. This is what Hardt and Negri (2004), regarding the concept of 'general intellect', call 'immaterial work', that is, 'labor that produces immaterial products, such as information, knowledge, ideas, images, relationships, and affects' (p. 65). For them, this represents a tendency towards an essentially qualitative and not quantitative transformation of work and social relations, in which 'the qualities and characteristics of immaterial production are tending to transform the other forms of labor and indeed society as a whole' (p. 65). According to this idea, it is necessary to recognize that the economic production that takes place in the contemporary metropolis implies the production of certain forms of life, subjectivities, practices and social relationseven if this occurs in an antagonistic way and always in a permanent constitution process.
Following the antagonistic perspective of autonomism and the relational vision of power proposed by Foucault, the contribution of this perspective lies not only in the offered diagnosis but also in how it is framed with contemporary social and political dynamics and processesfrom the question of technology to the forms of economic production. Hence, what is described above should not be understood as a Machiavellian plan to control society or as a total social relationship able to determine all social aspects of life. Rather, the authors highlight the existence of an antagonism where both relationships of capture and of resistance are present and become possible. That is, it is considered that, even from this set of transformations, other possibilities can emergewhile constituting an extension of production in society, there is also an expansion of the possibilities for conflict and creation of something new in the whole social field.

FRAGMENT ON MACHINES
Most of these ideas are based on an interpretation of Marx's chapter 'Fragment on Machines', in Grundrisse (1858). In a very succinct way, Marx's thesis defends the existence of a tendency in the development of capitalism of using technology, which represents a transformation in terms of surplus-value. The introduction and use of technology in the productive process would lead to a tendency towards a crisis concerning the law of valuea thesis that defends the existence of a law in the functioning of a capitalist economy that leads to a downward trend in the profit rateas well as the emergence of the hegemony of forms concerning relative surplus-value over forms regarding absolute surplus-value.
The shift, according to Marxist's theory, from the hegemony of forms of absolute surplus-value to forms of relative surplus-value suggests that an increase in surplus-value would essentially be realized through the introduction of innovationsas could be the case of technological innovation and not so much through an increase of labour (as was the case of absolute surplus-value). This transformation is associated with the change from ways of formal subsumption to ways of real subsumption of society by capital, in which life and society tend to be fully integrated in the functioning and reproduction of capitalism. As Jason Read states: '[r]eal subsumption is the transformation of the technical and social conditions of the labour process. A transformation in which what is originally outside of capital, the social and technical condition of labour, becomes internalized' (Read, 2003, p. 121).
This hypothesis is explored by Hardt and Negri (2004), who underline that it becomes particularly visible during the current phase of capitalism, given that various dimensions of life have come to be incorporated in the value production process and, therefore, qualitatively alter the very ontology of contemporary work. According to the authors, in contemporary capitalism society is 'put to work' and becomes productive itselfwith dimensions such as subjectivity, language, relationships, affections being incorporated into the productive process. As Read synthesizes: Capital no longer simply exploits labor, understood as the physical capacity to transform objects, but puts to work the capacities to create and communicate that traverse social relations. It is possible to say that with real subsumption capital has no outside, there is no relationship that cannot be transformed into a commodity, but at the same time capital is nothing but outside, production takes place outside of the factory and the firm, in various social relationships. (Read, 2009, p. 33) Nevertheless, and as stated by Matteo Pasquinelli (Pasquinelli, 2013(Pasquinelli, , 2015b, it is important to remember that, in Marxist terms, machines do not produce surplus-value. In this process, what happens is that knowledge (the 'general intellect' spread throughout society) is transformed into fixed capital through existing forms of technological mediation (Marx, 1858;Pasquinelli, 2013Pasquinelli, , 2015b. That is, social knowledge, not reducible to a single person and corresponding to a dimension considered immaterial, becomes technologically 'objectified' (in its diverse forms and expressions, such as in the production and development of algorithms) for the organization and production of surplus-value. Therefore, accompanying the theories here presented, it could be said that knowledge is appropriated and privatized within the dynamics of capitalism's functioning. According to Marx, this is also a process that involves a structuring of society for the production of value (Marx, 1858, p. 706). To conclude, it could be said that what is at stake in conjugating such contributions is the consideration that a transformation of capitalism implies, from the outset, a change and transformation regarding forms of life, subjectivities and ways of social organization in order to ensure its own reproductioneven if this always involves an antagonistic and never totalizing dimension.

GOVERNMENTALITY, METROPOLIS AND TECHNOLOGY
Algorithmic governmentality and metropolis The concept of metropolis will be worked here essentially using the reflections that result from the autonomist theory, even if they happen to be linked to some reflections of the theoretical body of Geographyfor example, with the work of Allen Scott (Scott, 2011). The autonomist thesis emerges not only from theoretical work but also from the observations concerning the spread of production and social conflict in society, giving special emphasis to the space of the metropolis. From that emerges the hypothesis that 'the metropolis is for the multitude what the factory was for the working class' (Hardt & Negri, 2009, p. 250). The metropolis is considered to be a space of antagonism par excellence, where processes of economic, social and political production occur. The metropolis is a space where technologies of power and possibilities of resistance and experimentation are in permanent conflictwithout power being able to acquire a totalizing dimension. The autonomist perspective highlights the antagonistic dimension of this process, pointing out the existence and importance of a cooperative and common dimension of economic production, and how that process produces specific social forms and relationships that are subordinated to the end of value production. Hardt and Negri stress that this production, even when collectively carried out, is captured by capitalism as rent, forming a situation of exploitation and diminished potential of such relations, and of the production and sharing processes that take place in the metropolis (p. 256).
For Stuart Eldon (Elden, 2007), there is a link between the organization of the territory and the three models of Sovereignty-Discipline-Security introduced by Foucault. Focusing on the currently hegemonic technology of powerreferring to the paradigm of Securityit is possible to find similarities between the analysis of the apparatus of power in relation to the territory and the questions invoked in autonomist literature regarding the metropolis. It stands out that the way this technology of power concerns the territory is not so much a strategy of limitation and contention, of raising barriersas occurred hegemonically in the paradigm of Disciplinebut rather one of minimum management and intervention of flow and circulation, aiming to stimulate such dynamics and allow actions and events to occur (p. 565). This rationale is associated with the neo-liberal ethos as identified by Foucault (2008) and with a conception of government as the management of effects (Agamben, 2015(Agamben, , 2016. This notion means the government would be focused not directly on the causality of problems but rather on the immanent management of their effectswithout necessarily having a solution in mind. For Agamben, security mechanisms are a permanent technology of government, essential to establishing and maintaining a given 'normality'. The concept of algorithmic governmentality is particularly useful for understanding the phenomenon and possibilities in question (Rodrigues, 2016;Rouvroy, 2011). Antoinette Rouvroy (Rouvroy, 2011, p. 122) proposed the concept of algorithmic governmentality to highlight how subjects and the physical world are 'taken as visible, endowed with meaning, assessed and produced' by automatic systems. From this notion a new episteme emerges, created in an immanent and relational way through the collection and analysis of data and information. This episteme involves: (1) the detection of the normal, that is, the detection of standards and common profiles, with the purpose of intervening in social practices and relationships in order to structure the possibilities of action; and (2) the detection of the 'abnormal', of something or someone that does not conform to and/or deviates from the common standard previously detected and, as such, must be subject to greater attention and control (Pasquinelli, 2015a;Rodrigues, 2016;Rouvroy, 2011). In the episteme's first form, when detecting the normal, the observed technology of power is associated with the paradigm of security, whereas when detecting the 'abnormal', the aforementioned technology of power is associated instead with the paradigm of disciplinethere is, therefore, a conjugation of both paradigms. Such forms of governmentality act towards applying a 'conduct of conduct', ways of acting, considered or calculated to a greater or lesser extent, which are designated to act on the possibilities of action of other people, according to a conception of government focused on 'how we conduct ourselves, how we attempt to conduct others, and how others attempt to control our conduct' (Rouvroy, 2011, pp. 7-8).

SMART CITIES AND GOVERNMENTALITY
There is a growing line of research that combines the study of smart cities with studies on governmentality. To exemplify, Vanolo (2014) highlights the emergence of a 'Smartmentality' as the new normative view of what a 'good city' should be, thereby influencing knowledge and practices of urban planning. Gabrys (2014) uses the concept of environmentality in order to analyse how 'citizenship is articulated environmentally through the distribution and feedback of monitoring and urban data practices, rather than as an individual subject to be governed' (p. 32). These two perspectives are related to the reasoning of this paper. In an attempt to present the link between the concept of governmentality and the phenomenon of smart cities, the works of Söderström et al. (2014), Kuecker and Hartley (2019) and Kitchin et al. (2017) also stand out, with the first two focusing on the issue of normalization associated with the emerging possibilities and effects of smart cities. This paper is more in line with the ideas presented by Kitchin et al. (2017) on governmentality. As the authors refer, and in line with Foucault's claim on the Sovereignty-Discipline-Security relationship, there is more an overlapping than a total replacement between technologies of power.
It could be said that these forms of governmentality, when associated with the assumptions and forms of intervention of currently hegemonic smart cities, take the urban space (including its population and subjects) as subject to analysis and management, by monitoring and programming the urban environment (Foucault, 2007;Gabrys, 2014;Kitchin, 2014;Kitchin et al., 2017). The possibilities of knowledge and intervention associated with the use of diverse digital technologies allow for the creation of databases with geographical patterns as well as a time series referring to various dimensions and territorial contexts, updated in real time (Gabrys, 2014;Kitchin, 2014;Kitchin et al., 2017;Scannell, 2015). Therefore, the epistemes created from such databases, generally employing a mix of diverse data and information, allow for the subsequent production of algorithms and applications that potentiate a more efficient analysis and management of the urban space. Said management implies the production of specified urban spaces and environments, populations and subjects, as well as structuring their conduct (Gabrys, 2014;Scannell, 2015). However, this does not constitute a situation of total control and impossibility of resistance.

SMART CITIES AND METROPOLIS
This paper defends the hypothesis that the contemporary metropolis, considering the state of digital ubiquity and the set of technological and socioeconomic transformations it has been subject to, relates and significantly updates autonomist theses about the metropolis. As a result, it is possible to establish the existence of a more significant blurring of oppositions, such as physical and virtual spaces, or between spaces and times of work and non-work. Given that several mechanisms currently allow for new forms of mediating and/or collecting and sharing data and information regarding various dimensions of life, the possibilities of bio-political production and real subsumption are intensified and the metropolis is completely transformed into a place of work in all its scale. Negri (2018) states that in these circumstances, and highlighting the potentialities associated with automation and the emergence of forms of control through algorithms, there is 'a higher level of real subsumption' (p. 167).
However, it is important to recognize that such a situation is always dependent on the context of its implementation and emergence (Rossi, 2016), and that it does not mean it will become absolute in terms of power relations. This phenomenon does not represent a state of total control or means it is free from any potentially emancipatory perspective, and it is not experienced in the same way by the populationrather, it is felt in a differentiated and unequal way. Therefore, and even if we consider that the whole population contributes to a collectively carried out bio-political production, and considered it to be essential to the current processes of innovation and economic production, it is possible to speak of 'inclusion by exclusion' (Agamben, 2014(Agamben, , 2015, a situation in which, due to the real subsumption of work, the metropolis and life as a whole are integrated in the processes of production of value. However, that might occur in an unequal and not typified way, by means of a contracted and remunerated relationship. As stated by Negri (2018): In the productive metropolis, disjunctive inclusion is understood as that which includes the entire population of the metropolisregarded as a productive spaceand distributes it therein according to mobile and flexible functions that are basically precarious in the construction of value … . (p. 98) Finally, it is important to highlight a critical aspect underlined by Pier Aureli (Aureli, 2017). The author also emphasizes that the history of the factory cannot be understood as a delimited space, but rather as a set of relations between various social spaces, flows and processes, in permanent transformation. That is, for Aureli, the metropolitan space, at least since the beginning of capitalism, could be considered as a space that expands the extraction of surplus-value from practices and social relations taking place there. As follows, this would not be exclusive of contemporaneityeven though it might be more intense today. Basing his theory on Tronti's (1976) findings, and highlighting the precursor and constitutive dimension of social struggle within the dynamics of capitalist development, Aureli stresses that the historical transition from the hegemony of absolute surplus-value to relative surplus-value (even if subject to different phases) implied the integration of all moments of life within capitalist production. Therefore, and even highlighting the specificity of the processes and dynamics that currently occur in the metropolis and contemporary society, the author emphasizes that the structuring of life and society according to the objectives of production and reproduction of capitalism is not exclusive to the present day.

METROPOLIS AND URBAN EXPERIMENTATION
In a rather schematic way, and given the processes of digital mediation and ubiquity, it is possible to approach the effect of such transformations on the metropolis from two different angles, considering both their forms of intervention and production. A first approach focuses on the emergence of new digitally mediated business modelsexamples being Sharing Economy, Open Innovation, Living Lab or Big Datawhich, while not being exclusive to or totally dependent on urban space for their development, find in it the ideal room for development, given their dimension and complexity (Rodrigues et al., 2015). The fundamental idea implies recognizing the metropolis as an enormous entity that produces data, information, algorithms and applications, with possible effects on the co-creation and production of knowledge and innovation (Chesbrough, 2006;Schaffers et al., 2011;URBACT II, 2015). This problematization is related to a line of research in economic geography concerning the territorialization of economy and innovation, as well as the prominence given by Hardt and Negri (2004, p. 249) on how economic production in the contemporary metropolis is dependent on the exchange of ideas, communication, knowledge and relational dynamics in general.
Another approach concerns the possibility, regarding digital ubiquity and mediation, of the metropolis becoming, in its entire scale, a space of production and work. This transformation refers not only to a possible reconfiguration of the working and non-working time and space, but also to the ways of value production within current society and economy, that can be extended to everyday life, which is taken as a decisive extension of value creation in contemporary capitalism (Hardt & Negri, 2009;Laing, 2013Laing, , 2014Negri, 2018;Scott, 2011). This particular line of thinking is directly associated with some of the phenomena previously presentednotably the processes of Open Innovation and Urban Living Lab that occur in contemporary urban space (Carvalho & Vale, 2019)but with a greater emphasis on the processes of urban experimentation related to 'test-bedding urbanism' or 'urban experimentation' (Cugurullo, 2018;Evans et al., 2016;Halpern et al., 2013), as well as the territory's infra-structuring associated with these interventions (Easterling, 2014). According to Evans et al. (2016), urban experimentation tends to be taken as a potential and alternative instrument in terms of urban policy. Part of this potential is related to the possibilities of experimentation and integration of populations in real projects of urban intervention, thus acquiring a potentially transformative character (Evans et al., 2016). However, Evans et al. (2016) also stress the need to consider the context in which such experiments are conducted in order to avoid the reproduction of unequal power relations. In those cases, the metropolis, or at least a limited area of it, is taken as a space of experimentation in a real context, where diverse technologies and solutions can be testedsimilar to what was identified by Gabrys (2014) concerning the programming of urban space and the environment, something that might imply as well as reproduce asymmetrical power relations. This paper argues that this aspect is an essential dimension of today's smart cities intervention logic, which will be explored in depth in the following section by analysing the See.Sense case.

ANALYTICAL FRAMEWORK
As follows, we propose an analytical framework with the following dimensions: Governmentality; Urban Technologies and Metropolis; and Urban Experimentation ( Table 1). The proposed innovative framework allows for a critical analysis of this topic by conjugating different bodies of literature that have not yet been integrated. For all these dimensions, a critical analysis of the ways of intervention and mediation of urban space and of society in general is sought. Like so, the great advantage and novelty of this analytical framework is its relational and critical dimension, having to also consider the diversity of urban phenomena prone to being analysed.
The theoretical review and discussion presented so far serve as a theoretical basis for such dimensions. With the dimension of Governmentality, we underline and analyse the way in which certain economic, technological, and urban interventions and transformations are directly associated with social changes and the structuring of the sphere of life. More specifically, it is important to consider the relations between technologies of power and technologies of the self something which is also present in the autonomist perspective, particularly in the way it highlights the antagonistic dimension of social processes and relations. With the dimension Urban Technologies and Metropolis, we analyse the introduction and use of several technologies in the contemporary metropolitan space in order to understand their scope in the mediation and structuring of such space. Thus, it is important to consider said introduction and mediation of technologies in the urban environment, emphasizing how they turn into a central entity within urban space in all its different dimensionsespecially given the situation of digital ubiquity and the possibilities that emerge with it. In like manner, it is also important to consider the antagonistic and asymmetrical power relations that can be verified in such processes, particularly at the level of production and transformation of urban and metropolitan spacewhether economic, urban, subjectivities and life forms. Finally, the Urban Experimentation dimension analyses how the metropolis, or certain urban spaces, are the target of experimental processes in a real context and at various levels. That is, it is necessary to consider how urban space is taken as a 'test bed', especially given the current technological possibilities, and how it is combined with its context and urban policies that may be associated with it. For all the above-mentioned dimensions, there is a critical perspective of possible arguments on instrumentality regarding the intervention and use of urban space. Such forms may be due to economic, technological or urban processesconsidering that, given their multidimensionality and centrality, contemporary metropolitan space concentrates and integrates these processes and phenomena relationally. For instance, it is important to emphasize the links between the subcategories 'Technologies of power', 'Power relations and structuring of the urban and metropolitan space' and 'Urban space as test bed'. Sense is a technological company specializing in urban mobility that especially focuses on cycling. The company has projects in other cities (also in collaboration with their local authorities) and sells its products in more than 50 countries. One of these products, See.Sense ICON, is presented by the company as a 'bike light' that reacts to the environment and uses 'advanced sensor technology' in order to detect 'changes of speed', being able to 'flashing brighter and faster at roundabouts, road junctions and filtering in traffic'. 5 The product can be connected to the cloud through an app. As is clear from the description, the product allows for the collection of automatic and georeferenced data and information, referring to the various dimensions of urban space and its users (but with guarantees of anonymity). 6 See.Sense ICON is the product implemented in the above-mentioned project. See.Sense was awarded one of the 'Open Challenges' prizes promoted by the Smart Dublin institutionan 'Open Challenge' related to urban mobility and aimed at encouraging bicycle use in Dublin. 7 With this purpose in mind, the See.Sense project was implemented. Smart Dublin is the entity responsible for coordinating smart cities' projects in Dublin's four local authorities, such as Dublinked, Dublin Dashboard and Smart Docklands. However, it is an entity with a marginal position within the structure of Dublin's local authorities. It should be noted that this is a structure with little range and resources (both human and financial). Thus, its main role is to mediate and articulate various interests . This makes Smart Dublin very dependent on Dublin's private sector initiative, which has a large diversity of projects and companieshowever, it is necessary to note that Dublin serves as the basis for the European headquarters of most of the world's major technology and digital companies. This is in line with Dublin's general urban policy guidelines. As pointed out by several authors, the orientation of Irish urban policies, and of Dublin in particular, was influenced by a neoliberal orientation (Kitchin et al., 2012;MacLaren et al., 2014). Since the 1990s, the principles of the socalled urban entrepreneurship, with the participation of private actors and a market orientation, have shaped the definition of policies, in accordance with a shift that has also taken place economically. As Kicthin et al. (2012Kicthin et al. ( , p. 1302 point out: 'In the 1990s Ireland embraced deregulation, entrepreneurial freedoms, and free-market principles and aggressively courted high-valued-added export-oriented FDI [foreign direct investment] (Kitchin et al. (2012(Kitchin et al. ( , p. 1302).' This model of development was questioned due to the urban, economic and financial crisis that occurred in Ireland after the crisis that emerged in 2007-08 (Kitchin et al., 2012). However, most of its elements have remained the guiding principle of public policy in both Dublin and all Ireland. It may be noted that smart cities' policies currently being implemented in Dublin fit into this strategic orientation and are taken as an urban and economic response to such a crisis.
As mentioned in the interview with a member of the company, the SeeSense project consisted of providing 500 'bike lights', which were used by employees of Dublin City Council and by workers of technological and cultural companies in the Docklands district. 8 The Docklands district makes for the current technological/digital and economic centre of Dublin, having undergone a major process of urban regeneration and transformation in recent yearsas such, one can guess the purpose for which this product is used in the area. We stress that, according to Heaphy and Pétercsák (2018), there is a long history of transformations and interventions in this district. Recently, another project called Smart Docklands 9 was announced, aiming to turn the district into a major urban experimentation for Internet of Things (IoT) and test-bedding urbanism. Promoted by Smart Dublin and Connect, a research centre, this project aims, among other things, to transform Docklands into 'the world's most connected business & living district'. 10 See.Sense ICON: an analysis Governmentality These projects and urban interventions structure urban space and its environment, endowing it with the technological means necessary for a situation of digital ubiquity, and allowing urban experimentations to take place there. Out of this scenario emerges a structuring of actions and subjectivitiesas could be the case for the cyclists involved in this project.
Cycling is a common practice for thousands of people in Dublin, each displaying immaterial and subjective dimensions. Thus, cycling is clearly digitally represented and mediated for diverse purposes. Cycling is something that could be extremely varied and is not completely reducible to external objectives and/or actors. However, this is an experiment for the cyclists themselves, who are, to a degree, voluntarily using both this product and the bicycle in their daily lives, and a situation of possible instrumental use by See.Sense and the local authorities in relation to them. That is, there is a relation between technology of the self and technology of powera situation of governmentality.
This was observed through a variety of data that was collected on the cyclists and their practices, as well as on traffic and the urban environment in general. The use of the See.Sense ICON product by the cyclists, particularly the data and knowledge emerging from said use, allows both See.Sense to improve the development of its product, and the local government to define new urban policies. This knowledge may alter the behaviour of cyclists, given their interaction with it. 11 And it may also alter the urban space itself, a change that is due, not only because it is necessary for present use, but also for possible future uses structured around the knowledge obtained from the experiment. This project enabled See.Sense, among other things, to identify the primary mobility patterns, and some of the main difficulties that cyclists faced on the road when studying their cycling behaviourit becomes clear, therefore, that this goes in line with the notion of paradigm of security, encouraging and allowing subjects to act, in this case, by giving their data. Furthermore, and given the relevance of accessing subjects' subjectivity, as well as the various dimensions of life that go beyond workplace relationships, the autonomist analyses of the contemporary economy and metropolis have obvious relevance. It may be considered that the forms of value production and urban experimentation analysed would not be possible without introducing the various spheres of life in the forms of economic, technological and urban mediation and production. This situation corresponds to a structuring of possibilities carried out through different forms of economic and algorithmic governmentality and an intervention in the territory. This is something that, even if not implying a totalizing situation, influences and mediates the experience of urban space for the population of Dublin, and the users of See.Sense ICON. The interventions, knowledge and effects created by both See.Sense and Dublin City Council may alter, even if slightly and not always in a necessarily negative and totalizing way, the practices of the individuals who participated in this project, as well as other cyclists and citizens in Dublin, and even other users of See.Sense product in different places.

URBAN TECHNOLOGIES AND METROPOLIS
This case is an example of a process in which the metropolis could be thought of as a space for mediation, production and even experimentation, enhanced by interventions, by the territory's infrastructure and by digital ubiquitythe sensors in See.Sense ICON are mobilized through bicycle use, and the various technological mechanisms mediating the urban space of Dublin. See.-Sense ICON, as a product that allows for the collection and production of data and reacts as well as influences urban environment, could be seen as an object capable of co-constituting and mediating the relations between human and urban space.
As already stated, such a process also reveals the importance, in a context where digital technology becomes ubiquitous, of accessing the subjectivity of subjects, meaning their incorporated knowledge as a potentiality for innovation. In this case, this is shown through the cyclist's knowledge in relation to his or her use of the bicycle and of urban space. This is something that exemplifies the emergence of digitally mediated processes of co-creation and innovation which extend beyond a situation of traditionally typified work relations. Subjects are considered to be the producers of data concerning various dimensionsurban space, everyday life, and even the See.Sense ICON product itself. This is clear with the recent release of See.Sense ICON2, 12 as well as with the usage of this new product in a similar project that also took place in Dublin, besides Manchester and Antwerp. The new project, SynchroniCity, 13 makes use of the possibilities available in the new product, and was able to collect a bigger variety of datasuch as locations, road condition aspects, mobility patterns, braking situations, collisions and near-miss events, speed and cyclists' profiles, among others. This highlights the importance of subjectivity, social practices and relationships for economic production, technological innovation and knowledge production for public policies. A knowledge created from a common dimension, related to spheres of life and of the metropolis, which could be used for multiple endssometimes without the consciousness of the subjects, and even may be used for processes that may be antagonistic to them. However, even if not forming a totalizing situation, reducible to an instrumental dimension, this is a process in which users may not be completely aware of their contribution to innovation and value production.

URBAN EXPERIMENTATION
As already stated, this introduces a logic of experimentation at various levels. First, regarding the product, the use of See.Sense ICON by cyclists allowed See.Sense to test the quality of and response to this bike light in real time. See.Sense accompanied the cyclists who participated in the project. As established by one of the heads of See.Sense in an interview, 14 that is the fundamental difference between this project and the retail sale of the See.Sense ICON product. See.-Sense collected the cyclists' feedback as well as the data regarding their routes using multiple variablestime, speed, traffic, roads, environment, etc. This does not happen with retail sales of See.Sense ICON, but with experimental projects, such as the case we are currently studying. The collection of such data and information is essential for understanding the enormity and complexity of data that this product, or others like it, may collect in the future, in other cities and in different contexts. As a result, this project of urban experimentation is also a way of optimizing the See.Sense ICON product and its potential, that is, a case of knowledge production and technological innovation, something visible in the release of See.Sense ICON2.
Policies such as the 'Open Challenges' and projects such as SmartDocklands, given their rationales and forms of intervention, consider and use urban space as a space for experimentation and production (of knowledge, value, etc.). The introduction of 500 bike lights in Dublin's urban space, even with the objects in question having a mobile dimension, structures the territory and influences its experience and mediationespecially in the Docklands district. The urban space itself, and particularly the areas considered more important in economic and technological termsas is the case of Docklandsis considered to be a suitable environment to test new products, as well as an entity open to intervention and knowledge production in a potentially instrumental way. This is also an example of urban experimentation regarding urban policy. The logic of 'Open Challenges' is one of experimentation, in which the public administration launches a tender so that private companies can respond to problems identified by the local authorities. In this case, See.Sense presented a solution for a previously identified problem, and that solution was tested in a real context. The benefits for Dublin's City Council, the local authority directly associated with the project, correspond essentially to the possibilities of the knowledge generated and potentially used for urban planning. Highlighted among these possibilities are: the creation of 'heat maps' and other types of knowledge regarding urban mobility patterns; the collection of miscellaneous data about the urban environment; and the monitoring of the state and quality of road surfaces and cycle-tracks. Some of the results of this project can be found on the webpage of the project. 15 Among several aspects, the results indicate the main mobility patterns concerning the use of the bicycle, the time periods in which the bicycle is most used, and the identification of sections of road surfaces in poor condition. In this sense, it could also be noted that this case illustrates an attempt to include the population in urban policies and intervention projects, even if the ends and ways of such participations are criticized.

URBAN EXPERIMENTATION IN DUBLIN: SYNTHESIS
It can be said that the introduction of the See.Sense ICON product in Dublin's urban space, particularly in areas considered to be economically, technologically and digitally more developed (as is the case of Docklands), is an example of two things: the intensification of forms of algorithmic governmentality applied to the territory, and the validation of autonomist theses related to the metropolis. The intervention and transformation of Dublin's urban space involved the introduction of various mobile sensors placed on bicycles as well as the dissemination of various forms of digital ubiquity. This project serves as an example for urban experimentation, as well as an example of an actually existing smart city (Shelton et al., 2015). The See.Sense ICON project allowed us to point out some of the limitations and potentialities of experimenting new solutions in real contexts. On the one hand, it allowed for the creation of knowledge used for the development of a product and for a better understanding of the dynamics of mobility in Dublin, which are essential to defining new public policies. On the other hand, the project reveals some of the difficulties and contingencies of the public administration at the level of the management of the contemporary metropolis, and also lets us point out possible situations of instrumentalization and reproduction of asymmetric power relations when such experiments are carried out.
This case also enables us to ponder the evolution of the implementation of a smart city project, and its possible collateral effects, revealing the unpredictable and accidental nature of urban experimentation projects. Because of a project related to urban mobility and the use of bicycles in urban environments, some of the cyclists have become mobile data producersand their status as cyborgs could be questioned. Additionally, the city of Dublin, through other implemented projects, goes from an accidental smart city to an articulated smart city . As follows, something as simple as the use of the bicycle ends up becoming connected with external objectives, such as focusing on data production and product innovation, developing more Smart City projects and asserting Dublin as a Smart City, among others. However, this is not a totalizing power relation. Cyclists continue to use their bicycles for varied reasonssimply because it is a practical way to move through the urban environment, for ecological and/or health reasons, for the love of cycling, or the importance of aesthetic and symbolic elements mobilized by the individual as factors of social and cultural distinction. Moreover, the very aims and intentions of the actors who promoted this project could be questionedwith not anticipated data and forms of knowledge arising and other uses for the bicycle and of See.Sense ICON product, representing another type of change in urban environment. Given the complexity of the corporeal activity and the urban experience, Tironi and Valderrama (2018) state that there are more and more ontologies that are not and cannot be represented by data.

CONCLUSIONS
In this paper, we argued that both the currently hegemonic transformations concerning smart cities and the forms of technological and digital mediation in the contemporary metropolis represent a form of governmentality that is, at the same time, economic and algorithmic. Therefore, we maintain that such interventions in urban space tend to have as their objective not only a form of economic production, but also a transformation and structuring of territories, populations and subjects. The introduction of the ICON See.Sense product in Dublin is an example of how certain types of urban experimentations, given their logic and technologies of power, could constitute a form of governmentality. And even though this is not exclusive to the aforementioned phenomenon, it is argued that, given the situation of digital ubiquity and the diverse forms of digital and technological mediation in the contemporary metropolis, such interventions represent an updating and intensification of the autonomist theses related to the metropolis and the forms of bio-political production occurring in it.
Thus, we consider that this theoretical discussion, arising from diverse perspectives, forms a useful and relevant reference to the construction of a critical and comprehensive analytical framework. While recognizing that there are several critical analyses of smart cities, we believe that the one presented here could not only improve the understanding of the phenomena directly associated with technological and digital transformations, but could also frame them in a set of more general phenomena and processes, such as those related to technological mediation and to forms of contemporary economic and urban production. This perspective allows us to synthesize and connect various social, economic and territorial changes, which grants a deeper understanding of such phenomena.
The analysis of the See.sense ICON project, in its various aspects, allowed for a critical and global understanding of the urban, social and economic phenomena discussed above. Said analysis showed how a variety of processes was combined and how it contributed to the transformation of urban space, to mediating and structuring the practices and social relations of certain subjects, as well as to the technological innovation and economic production of See.Sense ICON.
However, in future research, it would be important to explore in depth the specificity and ontology of urban space within processes of technological and economic mediation and transformation. Assuming there would be a change in urban ontology, it would do to explore the machinic and infrastructural dimension of what is specifically urban in terms of the mediation and the structuring of life and society within the contemporary metropolis.
One other thing that could be done in the future would be to not strictly follow the tendency that is characteristic of autonomist analyses (Negri, 2018), and, instead, study the socio-spatial inequalities associated to such phenomena. Recognizing that people's day-to-day, in terms of bio-political production in the metropolis, is unequally mediated and experienced, it could be relevant to apply a different type of methodology to address said inequalities. Perhaps, such an analysis could confirm the hypothesis that urban centres are more technologically and digitally mediated than certain urban peripheries, or that there are differences in the possibilities of use by different subjects.
Equally relevant for future investigations is the problematization of proposals for different uses of technology and contemporary urban spaces. It could be stated that the main challenge is in maintaining most of the technological innovations and mediations currently used for governmentality purposes, without them giving rise to situations of control, to the structuring of possibilities and the reproduction of inequalities. It is a question of the relationship between technologies of power and technologies of the self as identified in this paper, as well as the wide-ranging debate about the neutrality of technologyeven if assuming, from the outset, that technology mediates and produces effects. A possible research hypothesis could explore the potential of an 'ethical-political use' (Agamben, 2016) of technology, considering that such reflection implies forms of subjectivation and creation of forms of life that are critical to situations of instrumentality and exteriority in relation to the self. It would also be important to identify and analyse current forms of use of smart cities and forms of digital mediation in the contemporary metropolis that could go in a contra-hegemonic direction to the one identified, thereby highlighting the possibilities of resistance currently found in such situations. For example, and referring to the case examined above, it could be interesting to study the uses of the bicycle or of the urban space itself. Also, studying the forms of digital mediation of See.Sense ICON that are not in line with the rationales and purposes anticipated by the actors involved in the project could also be noteworthy.