Navigation – Plan du site

AccueilNuméros12DossierGoverning the City: The Detecta S...


Governing the City: The Detecta Surveillance System in São Paulo and the Role of Private Vigilantism in the Public Security

Alcides Eduardo dos Reis Peron et Marcos César Alvarez


En 2014, le gouvernement de l'État de São Paulo a annoncé la mise en œuvre d'un système de surveillance qui permettrait l'identification automatique des criminels et la synchronisation de plusieurs sources de données afin d'améliorer l'activité policière. Cependant, depuis le déploiement de ce système censé contribuer à réduire la criminalité et d'autres formes d'arbitraire, les brutalités policières sont demeurées problématiques. Le modèle de dispersion des caméras dans la ville de São Paulo signale l'émergence de pratiques de sécurité centrées sur le gouvernement du malaise et la défense des intérêts privés. À partir de l'analyse de la gouvernementalité chez Foucault et des “Critical Security Studies”, nous examinerons la manière dont l'introduction du système Detecta engendra une forme hétérogène de maintien de l’ordre, axée sur le gouvernement de la « qualité de vie » dans les périmètres de sécurité et le partage des responsabilités avec les organisations privées. Il semblerait selon nous que ce processus renforce les pratiques traditionnelles de discrimination, en créant une atmosphère de méfiance permanente, brouillant les frontières entre le public et le privé dans l'appareil de sécurité.

Haut de page

Texte intégral


1In 2014, the São Paulo state government announced a partnership with Microsoft and the New York Police Department to import a state-of-the-art surveillance and monitoring system, baptized as Detecta. The original system, called Domain Awareness System (DAS) is characterized as a technology for tracking and profiling criminal and suspicious conduct from analytical videos, integrating criminal databases and police officers. This system was developed in New York years after the September 11 attacks as a way to mitigate the possibility of terrorist plots and increase police capacity to anticipate and react to criminal practices (NYPD, W/D). The peculiarity of these technologies is their capacity to analyze massive amount of public data, and analytical images, crossing with platforms of data criminal, and multi-order data classifying groups of individuals, issuing alerts if any programed action is detected.

2This way of managing security gains a new notation: predictive policing. The RAND Corporation defines the practice as: “Application of analytical techniques to identify potential targets for police intervention and crime prevention, or solution of past crimes from statistical predictions” (Perry et al., 2013, p. xiii). Microsoft argues that predictive policing derives from a new arrangement made possible by the new technologies of cloud analysis and storage, evidently considering the prospect of cost reduction coupled with greater efficiency of police actions in the “fight against crime” (Arthur, 2015). Moreover, to bring a more substantive understanding to the predictive turn in police practices, Harcourt (2007, p. 01) relies on the concept of actuarial methods in criminal law, using a definition that portraits it as “the use of statistical rather than clinical methods on large datasets to determine different levels of criminal offending associated with one or more group traits, in order (1) to predict past, present or future criminal behavior and (2) to administer a criminal justice outcome”.

3In this sense, the DAS possess a set of features to conduct crime analysis and is meant to further reduce the ‘siloing’ of relevant data, structuring the New York Police Department as a center for the production and dissemination of information and criminal trends. This "statistical-predictive" knowledge, characteristic of military doctrines for counter-insurgency activities in the “Global War on Terror”, now returns as an urgent response to domestic terrorism, which demands exceptional efforts to prevent its occurrence, including police pre-emptive actions. More than that, this same technology, invoked as an exceptional measure for the fight against terrorism, ends up having dual functionality, employed to combat the criminal activities, deviations and infractions of ordinary citizens, making them targets of suspicion, as Graham (2016) notes.

4However, companies such as Microsoft and Genetec have bundled these systems as "smart solutions", capable of edifying fast, connected, and secure cities worldwide. Moreover, these companies are responsible for stimulating a culture of surveillance on police and on the citizens, either offering technologies and helping police to reorganize their operationality, or empowering neighborhoods associations with technologies and know-how to acquire and sustain surveillance systems. One example of this is the Program Citiwise from the Genetec, which, under the mantle of the concept of “smart cities”, seeks to conciliate urbanism, security projects, private companies, and several of other institutions to enhance what they consider to be safety through interconnected surveillance systems. In São Paulo, the neighborhood Vila Madalena was the pioneer in adopting this projects, and according with some of those enrolled, the more the region is crowded with sensors and cameras, the more security would be enhanced: “I’ve seen a huge development in use of sensors, from sensors that measure the quality of water to sensors that turn a camera on when they detect the sound of a pistol that has been fired. The next step in a project like São Paulo is to take all this real-time information that will be collected and identify the patterns that occur, enabling predictive analytics: when will things happen, what will the traffic be like at any given time, how will the weather impact city life and crime?” (Genetec, 2019).

5Embedded in this same rhetoric, public powers found a subterfuge to move on with public-private partnerships for the most diverse fields of action, bequeathing the citizen with the responsibility for promoting the surveillance apparatus. The discourses surrounding the Detecta system in São Paulo seem to follow this idea of smart and integrative solutions, involving technology, private surveillance systems, information and data, and citizen vigilantism as a way to guarantee security, and better equip the police authorities to act against crime (SCC, 2017). However, as we will present latter, the intentions in the deployment of Detecta are much more related to the organization of public private assemblages to govern public security, reducing the costs of its administration while splitting its organization with citizens and private companies, as well as to disseminate a policing organizational model centered on data administration and statistical analysis.

6Thus, drawing on Foucault’s analytic of government (Foucault, 2008), through his idea of dispositif, and on the Critical Security Studies, without neglecting debates related to surveillance, urbanism, we explore how the instauration of Detecta system enabled a heterogeneous form of policing, focused on governing the circulation of people in certain perimeters, and sharing responsibilities with private organizations in this process. Even though Detecta does not fully incorporate the "predictive" technologies, such as the development of data banks, and the video-analytic functions, our hypothesis is that far from overcoming the usual police arbitrariness (Mesquita Neto, 1999; Cardia et al., 2003; Adorno et al., 2008) , these surveillance systems seem to reinforce traditional practices of discrimination, creating an atmosphere of permanent mistrust, blurring and weakening the lines between public and private in the security dispositif. Thus, we believe that this reordering of the security dispositif, and the intensification of its permeability (to private and transnational interests), may lead to discretionary and discriminative practices, whenever biased watchers, citizens, and other non-state actors become central to perform security.

7Although this work mobilizes an extensive bibliography of the Surveillance Studies field, especially regarding the phase described by Castagnino (2018) as the understanding of surveillance as a fundamental element for the governmentality and ordering of the world, it does not only intend to address the existence of new forms of surveillance or make superficial accusations about the technical progress or the ills of new technologies. In fact, following Castagnino’s criticism (2018, p. 31), our efforts are to understand the “surveillance effects”, that is, how private companies, state, municipality use a form of surveillance that seems disproportionate to social needs to such an extent that they seem to disrespect the dignity of the person.

8Then, we relied on ethnographic research and interviews conducted between 2017 and 2019, from incursions at electronic security fairs, police congresses, visits to neighborhoods, as well as consultation of documents and interviews with residents, businessmen and security agents. In these field activities, we sought to understand how these practitioners perceived security, crime, and how they mobilized a specific discourse to made up their point. Moreover, through visits and interviews in these security perimeters, we wanted to see how a new way to organize the vigilantism seem to support a permanent state of mistrust.

9This work is then divided as follows. Firstly, we wanted to explore how the idea of vigilantism that permeates this culture of control was developed and subsequently inscribed in DAS. Secondly, we address the process of adaptation of the Detecta in Brazil, exploring the discourses and practices that lead to the formation and stabilization of the security dispositif. Afterwards, we offer a brief explanation of the methodologies used in our research. At last, we describe the formation of the security perimeters, and how they become territories where segregationally practices seem to be reinforced.

Security as Control: From DAS to Detecta

10Between 1970 and 1990, several American cities begin to observe an increase in crime rates, whether in terms of violent crimes, or minor and patrimonial robberies. According to Wacquant (2003) and Garland (2008, this was related to a greater economic weakness and bankruptcy of a Social State, as well as to a series of legal punitive actions whose focus was on a type of crime practiced by the poorest segments of the population. In that time, crime management intertwined with the formation of a penal state, directed to the poorest layers, culminating in practices of mass incarceration of these communities.

11 It is only from the late 1980s onwards that the NYPD sees a reduction in crime rates, as Belli (2004), Parenti (1999) and Chronopoulos (2017) point out, not due to the series of criminal enforcement measures, such as the “broken windows” theory (explaining the fear of crime in terms of social disorder) in the 1990s, but due to a series of economic, political, and institutional factors, like the introduction of Community Police, the increase of the police budget from 1993 on, more investigation on police corruption, and mainly the economic growth and development of New York. These changes were mostly influenced by the emergence of a neoliberal rationality – utilitarian, based on the theories of rational choice, and often focused on an optimum management and equilibrium of resources – that impacted both the management of resources at the municipal and intra-office levels, and in the way police organized themselves and acted over crime – on one side structuring themselves as a center for production and diffusion of information (Haggerty and Ericsson, 1999), on the other, aiming at reducing incentives to crime, for example. This rationality was impacted by some processes that occurred parallelly, changing police practices to confront “illegalisms” in the United States, the emergence of a new police model focused on an environmental control of the cities (producing knowledge about the city, the movement and circulation of people, etc.), the development and application of information and communication technologies, making the police apparatus more efficient, and, finally, an intensification of the practices of “communitarian police”, a series of techniques that would bring police close to the community, enhancing its efficiency (Skolnick and Bayley, 2006). When Rudolph Giuliani assumes as mayor, in the mid of the 90s, elevating the spends with the NYPD, and supporting a series of internal reforms, all these processes and rationalities converged in what David Garland diagnosed as a “culture of control” in policing activities (Garland, 2008).

12 In the early 2000s, when an “advanced liberal” technology of government was being developed, Garland points to changes in the policing model in the United States, moving away from reactive strategies of direct facing crime. He demonstrates how a range of agencies, practices, discourses, and policies will redirect their efforts towards producing targeted and community-based forms of policing aimed at "safeguarding order" and policing “quality of life”. According to him, “policing became smarter” by approaching the community and emphasizing prevention by focusing on local circumstances for resolving crimes (Garland, 2008, p. 367-368).

  • 1 We realize that this diagnoses still open, and there are several critics on it. However, it fits as (...)
  • 2 At this point is fundamental to mention the work of Clarke (1980), who also explores how this drive (...)

13 Essentially, Garland diagnoses that a new culture of police control was in emergence in the United States1, in which “information technologies and new managerial techniques have combined to produce greater control of resources and more targeted and punctual conducts” (Garland, 2008, p. 368). This culture is manifested through the cooperation of several public and private actors in providing a kind of situational prevention2 resulted from environmental criminology. In addition, the adoption of cognitive assumptions linked to a neoliberal economic rationality is the mark of this new police culture, in which “the costs of crime are now routinely calculated, as are the costs of prevention, policing, prosecution and punishment; the numbers produced help guide policy choices and operational priorities” (Garland, 2008, p. 396). In Garland’s words:

Instead of concentrating on individual criminals, the preventive sector seeks criminogenic situations that may be altered to let them less propitiatory for criminal events, less inviting to potential criminals. It analyses the flows of people and the distribution of criminal events, identifying “hot spots”, products and recurrent patterns of victimization, transforming them in objects of action. (…) Proportional to the way that the government organizes, optimizes and directs the capacity to control the crimes of citizens, corporations and community, it simultaneously extends its governmental reach and transforms its mode of control (Garland, 2008, p. 371).

14 This diagnosis on the new culture of control has a strong parallel with the notion of governmentality explored by Foucault in the 1970s, where the dynamics of state action is substantially altered, figuring as a technique of government that relies on the population and instrumentalizes an economic knowledge as a way of producing and conducting adequate behavior. It presupposes a permanent administration of fear since the constant production and reproduction of threats to “freedom” are means of expanding the available instruments to combat and manage these same threats. It comes from the articulation between a Pastoral Power, which is anchored in the idea of salvation of the “flock” as a justification for conducting its conduct, and in the emergence of the liberal art of governing, which conceives the government as a reactive technique to the demands of a social body with an apparent economic-utilitarian rationality (Foucault, 2008, p. 298). In this context, government technologies operate in a way that systematizes, regulates and stabilizes social and power relations, avoiding both the dissolution of individual freedoms or the imposition of sovereign power and domination, as Lemke (2017, p. 27) points out. Thus, “governmental technologies bring together scientific knowledge, technical devices, anthropological hypotheses and architectural forms in strategic ways of establishing relations of conduct” (Opitz, 2011, p. 22).

15 The culture of control, therefore, is compatible with a governmental dynamic, since both have the objective to modulate the social body, either in its flows or in the managing of the risks to the system’s instability. This is what Garland's idea of a culture of control is, a kind of criminology of everyday life, in which the processes and social arrangements in which people are immersed in, need to be integrated to produce fewer incentives to crime, a kind of expanded control dynamics for social processes. As Garland points out: “(...) the criminology of daily life approaches the social order as a problem of system integration. They are no longer the people who need to be integrated, but the social processes and arrangements in which they live” (Garland, 2008, p. 388).

16 At the same time, the concept of community policing become systematized and widely adopted across the United States. It consists of a series of practices like decentralization of the police action and range, allowing more police units and stations focus on specific areas, getting close to the communities; the involvement with communities, in which citizens are stimulated to exchange information with the authorities, becoming an important asset for combating the crime, and finally, the introduction of specific methodologies to help police to and structure actions that may help to solve some community’s problems (Ribeiro, 2014, p. 432-534).

17 In face of these changes that the American policing systems begin to reorder, the adoption of monitoring and surveillance systems contribute enormously to the establishment of a regime of visibility whose center is the production of knowledge about society, the construction of heatmaps over regions where crime occurs, enabling some proactive police action over reactive ones. Among these systems, the Compstat (Computer Comparison Statistics) stood out as a planning tool, aimed to produce detailed criminal statistics and, forcing the necessity of agility and efficiency over the police staff (Belli, 2004). All these changes had the objective “to use the collaboration of the population and modern technologies to achieve scientific and neutral forms of risk control, through the construction of individual and social trajectories, demarcation of territory and borders between populations at risk and the others, and the analysis of the decision on who is really dangerous” (Bigo apud Belli 2004, p. 80).

18 This “culture of control” is generalized and sophisticated with the development of the DAS by Microsoft and the New York Police Department in 2009. It integrates information from diverse databases with analytical camera systems, and peripheral police devices, enabling greater efficiency of service activities and dispatch, the construction of statistics and heatmaps of criminal practices, and proactive police actions. It was vital for New York police and anti-terrorist agencies to adopt an instrument that would enable automatic identification of potential terrorist threats and criminal conduct. The DAS, in this sense, is portrayed as a tool of counterterrorism, but also able to contain demonstrations, and minor crimes (NYPD, 2009, p. 02).

19 Thus, the DAS reordered public security, introducing a system-wide and permanent vigilance and suspicious to govern the conducts of the general population, and also brings private companies and technologies to the sphere of public security, not as merely providers, but as actors with great capacity of agency in the system. This active private role over security is intense in New York since the 2001 terrorist attacks, and increased in the last ten years, as seen in the role of McKinsey consulting the NYPD, helping it to “shape its future” (Amoore, 2013), in the public and private partnerships promoted by the SHIELD program – a public-private partnership oriented to increase the exchange of information between private companies and government in issues related to security – , or in the Wall Street firms operating the Lower Manhattan Security Initiative (LMSI), where more than 3000 cameras were installed in Lower Manhattan in a public-private effort in 2007.

20In this sense, how this public-private, human and non-human arrangements are able to organize the security as a technology for control and government, sustaining and scattering a “culture of control”? To move on with these questions, we should address issues that have been raised by the Social studies of Science and Technology (STS) and Critical Security Studies in the last years.

21As proposed by Côte-Boucher et al. (2014, p. 196), security policies should be understood as a set of varied practices that maintains complex relationships with security discourses, legal and policing regimes, experts and technicians. Advocating for a “practical turn” in Critical Security Studies, which focus on ideas, policies, institutions, and actors in the fields where security is exercised, the authors argue that sustaining security as practice makes the concept lose its abstraction, and be understood as this assemblage of routines, technologies, and institutions. Security in its practical manifestation would be a translation process, where "policies are not only transmitted and implemented: they are translated into new genres for each step and each new actor or scale involved, and simultaneously, the processes generate new spaces of contestation" (Côte-Boucher et al., 2014, p. 1999).

22Similarly, Borne et al. (2015, p. 312) understand security as a practical set of actions, however, composed and intermediated by technologies, technical artifacts that carry within themselves a program of political action that will organize, change, and even re-signify security policies in practice. Therefore, the understanding of policing and security activities needs to be observed from the practice, verifying how concepts and abstract orientations are translated in the field. As Aradau (2010) also proposes, security policies are sociotechnical assemblages where discourses, practices, values conform to the process of materialization of technology and matter itself. In this sense, the idea of Borne et al. (2015) is precisely to point out that, if security policies are composed by technical objects in their practice, it is fundamental to recompose the process by which security demands instincts the understandings and the routine procedures were previously summarized into code lines and programs of action that resulted in these artifacts.

23 This perspective approaches to what Latour purposes as technical mediation. He understands that technology is “society made durable”, a set of relationships, programs of actions, inscribed in technical artifacts, bringing stability for the relations between humans, and displacing actions (Latour, 1991). Two senses of mediation are important for us here, the idea of translation, and the idea of composition. In his study, translation means “displacement, drift, invention, mediation, the creation of a link that did not existed before and that to some degree modifies two elements or agents” (Latour, 1994, p. 32). In short, the relationship between human and nonhuman actants displaces a previous course of action, creating “detours”, new courses and programs of action. The sense of composition means that a series of goals, steps, actions, and intentions are made possible through the assemblage between man and technology, as he points out: “(…) actants are in the process of exchanging competences, offering one another new possibilities, new goals, new functions” (Latour, 1994, p. 35).

24 It is in this sense that the culture of control is inscribed in the DAS. The idea of ubiquitous and permanent visibility, and also of constant suspicious are inscribed in the DAS through the code lines that compose its algorithms, and the infra-structure disposition of the system, making them “durable” and performative in society. This is clarified by the argument of an interviewed Microsoft executive – who actually was responsible for intermediate the relationship between the company, NYPD and São Paulo’s Public Security Secretariat – , "(...) [the DAS] has a price, from which is the source code that has been built on many good practices, and the translation of this knowledge is added to the reality of that new client. This is the concept called solution, where the context of the Detecta is inserted "(Microsoft Executive, 2018). Then, these technologies translate and compose the routine of police activity, into a more data-intensive practice, bringing up “awareness” as an operational and useful concept. Not only we see the launch of a new model of security expressed in the notion of permanent surveillance and suspicious, governing the conducts of people to produce stability and order, but we see that it is inscribed on a commodified apparatus, with the active agency of private transnational companies.

25Both DAS and Detecta are, then, the result of this public-private initiative to build and consolidate internationally a standard for “good” policing and surveillance practices, once the latter is an attempt to replicate through the design of this “solution” the dynamics of situational prevention and awareness and the culture of control in Brazil and other countries. The Detecta, as we will present latter, for specific reasons does not incorporate yet the same functions as the NYPD’s DAS, such as image’s analytics or facial recognition systems, it integrates images from all cameras connected with it (at municipal and state’s level), as well as data from several police agencies, such as criminal records, SP+Segura data, robbed car’s ocurrences, notifications of accidents, and other occurrences. On one hand, Detecta eases the process of investigation to the police, and seem to help them to gather a great amount of data related to a determinate occurrence, providing “Situational Awarenes”. On the other one, as we will present further, it presents a model of surveillance that is not specifically directed to combat the crime, but mainly to attend the demands of local wealth social groups, and also to constrain certain groups in degraded zones in São Paulo, scattering a kind of culture of control among the police forces.

Brief notes on the methodology

26From 2018 to the mid of 2019, we interviewed several security entrepreneurs, police technicians, officers, public authorities, users of the Detecta, leaders of associations, and also visited surveillance and monitoring centers, technology and defense fairs, as well as events of condominium security, and even attended in Vigilância Solidaria’s public meetings. We analyzed some official documents from the Secretariat of Public Security or other institutions, to comprehend how the system was portrayed, the negotiation about the technical details proceeded, the system was elected, and its actual condition. We also considered material diffused by the media, such as interviews, TV programs, campaigns, and quotes from the actors central to the development of Detecta. Our purpose was to understand how these actors perceived security and their place in the network of relations under development, and how these actors and documents portrayed surveillance as an immediate response to the dilemma of crime and disorder in the city.

27It may be a bit tricky to develop an approach on a sociotechnical system like Detecta, in first place because it was not properly mediatized by the Secretariat of Public Security, it is not fully operational, and fundamentally, it is not always easy to access decisionmakers, agents and documents – which are sometimes classified. Due to this, we decided to approach Detecta, first, from its marginalities, this is through some of the private agents and companies who operate private systems and programs related to Detecta, its users, and in symposiums, fairs and meetings promoted by private agents, neighborhood and commercial associations to divulge Detecta. Secondly, we approached public administrators responsible by the technical, operational and management of the Detecta and public security. Finally, we were able to talk to some users of the systems in neighborhood and commercial associations, visiting the COPOM (Center for Military Police Operations) which runs the Detecta system, and also take part into two meetings of the Vizinhança Solidaria initiative, which is a fundamental technology to get together police, residents, and sociotechnical systems. During this whole period, we were raising several primary and secondary documents and material, both produced by authorities, media and private security companies.

28During this first marginal approach, we intended to identify those who were central agents in this dispositif of public security, their role and importance in this context. We, then, visited around 8 electronic security fairs, panels and security symposiums in São Paulo, such as the ISC, LAAD, Exposecure, Simpósio da Segurança Condominial, Drone Show, Congresso Abese, Exposec 2018 e 2019. In this events, we mapped the main companies, entrepreneurs, we characterized them by their function and we collected printed materials, attended to conferences, identifying common speeches about security, surveillance and police efficiency, and also we were able to get the contact of private and public agents related with Detecta. When these contacts were established, we proceeded with interviews with executives and entrepreneurs of private security companies, and informational system companies. In this phase we interviewed around 5 individuals, two of them CEOs or executives from national electronic surveillance companies, one of them a CEO from a transnational electronic company, other one an executive from Microsoft related to the implementation of the DAS solution in Brazil, and the last one of them the president of the ABESE (Brazilian Association of Electronic Security System Companies). These interviews allowed us to understand how Detecta was conceived as a public private assemblage, and a way to bring private agents into the public security apparatus. Moreover, these agents helped us to understand how the system was expanding fundamentally over mid and upper class neighborhoods, mainly because these groups could afford the system. These neighborhoods, such as Morumbi, Alto de Pinheiros, Pinheiros, Higenópolis, Jardins, Planalto Paulista are huge neighborhoods, in some cases with hundreds of thousands of inhabitants, in process of gentrification – which means that in some cases, there are visible social contrasts – with commercial areas side by side with residential areas, great parks, gated condominiums, mansions and residential buildings. In a short walk around these neighborhoods, it is possible to notice the proliferation of plates from City Cameras and Vizinhança Solidária, as well as cameras in scattered among the facades of the houses, when not arranged on specific poles, pointing to every direction, dividing space with private security guards and informal security “watchers”.

29These interviews also helped us to identify public agents related to the implementation of Detecta, who we approached afterwards, interviewing and paying visits to the COPOM. In this phase, besides the visit at the COPOM, where we were received by the head of the installation – who also didn’t authorize to record the conversation – we interviewed three other public agents who were responsible for the implementation of Detecta, and its technical maintenance and deployment, the head of the security of a public of São Paulo, who brought the Detecta system to the university, the head of the Coordinator of the Group of Information Technology of the São Paulo’s Public Security Secretariat, and the São Paulo’s municipal secretary of public security. All of them were able to provide us with details of the history of Detecta, the way it is deployed, detailed how subsystems such as City Cameras and SP+Segura are connected to the system and help to expand its network of cameras and sensors.

30Finally, we were able to visit some spaces, and talk to people who directly operate systems and subsystems related to the Detecta. These visits allowed us to understand how human machine interactions were performed, how control rooms were elaborated, and these users’ expectative and perspectives about the camera systems, sensors and security. In this phase We visited two control rooms and monitoring centers, one in the Hospital das Clinicas (which is introducing cameras connected to Detecta, and even analytic images algorithms), and one public university’s control room. We also were able to talk to the head of the SAAP, a neighborhood association of Alto de Pinheiros, to understand how these residents organize themselves, together with the private companies and public security agents to perform security in their neighborhood. Moreover, several of these neighborhoods were visited, we circulated around it and registered the disposition of cameras and other sensors.

31Then, interviewing agents from the margins of Detecta, until its core, the administrators, operators and users – always through semi-structure interviews, to allow these agents to express themselves, and at a certain point to present new issues – we were able to easily circulate among these agents, to understand how they relate to each other, and how do they perform security and seek to stabilize the dispositif.

Towards a Security Dispositif in São Paulo?

32If we want to use the idea of dispositif to address São Paulo’s new security arrangements, we might focus on the Critical Security Studies approach, a specific field from Security Studies concerned with the “collateral effects” of security policies, in other words, how the performativity of security is inherent to the production of insecurity. This is to say that every security policy directed to some groups may marginalize, criminalize or even securitize other groups. The Critical security studies focus on the interaction of public and private agents (both national and transnational ones) in the performativity of security, on day by day security practices, understanding that security is ultimately manifested in these routines. Moreover, authors like Bigo (2008), Lemke (2017) and Optiz (2011) help us to problematize concepts like Foucault’s dispositif, mainly when connected to security practices, once consider fear and insecurity as fundamental elements to the stabilization of security dispositifs. Along this section, we seek not only to describe the formation of this security dispositif in São Paulo, but through the critical security studies, we try to understand how this process blurred the lines between public and private services, at the same time mobilized fear and insecurity as fundamental tools for its expansion.

33For the last ten years, officials from the São Paulo Military Police developed many studies referent to the necessity of a reform in the way the police manage its resources and organize its practices to combat crime. Examples of it are the studies developed by Military Police Lieutenant Manoel Mello (W/D), who sustains that a new police managerial system should involve “intelligent” technologies, capable of integrating data from various silos, and allow a more efficient model of command and control. Following some of these recommendations, in 2013, the Secretariat of Public Security created a commission to investigate new technological solutions for security purposes developed around the world. This commission visited London, New York, and Amsterdam, analyzing surveillance technologies, communication solutions, and proceedings adopted by the police departments. A report produced by it concluded that the DAS is the most advanced technology among the ones seen in the visited cities, and that the adoption of similar models would be decisive to produce “more effective [criminal] evidences” (Assessoria, 2013, p. 329).

34The idea of societal surveillance, centered in a control room, with automatized analytic cameras stimulated in the authorities an imaginary of efficiency only achieved by the adoption of this technology, and also helped to profuse an image of a modern and sophisticated government – this enthusiasm is perceived in the attachments at the end of the document, a summary of the positive repercussions of this visit in the Brazilian media (Assessoria, 2013). At the time of its adoption, the announcement of Detecta was made through propaganda in television, which framed a cyclist jumping off his motorcycle with his helmet on, and being automatically portrayed as a suspicious person. Right after it, the governor was framed been congratulated by the Microsoft´s vice president, for the astonishing monitoring center he had built, affirming that São Paulo had taken an important step to combat crime.

35In an effort to reduce costs, and administrative clashes in police institutions, as well as to be more efficient and to produce more visible effects on crime reduction, the Detecta would have both the functionalities of images analytics, as well as the production of heatmaps in areas of crime occurrence, making police activity more taxable and proactive. According to the Secretariat for Public Security (2015), the intention was to expand the range of suspect profiles in addition to traffic-related activities, by crossing information from databases of other institutions (Governo do Estado de São Paulo, 2015).

36However, since its adoption, just a few information regarding the operation of the Detecta system has been revealed. Moreover, in the year 2016, a report was produced by the Court of Accounts of the State of São Paulo, which states that the system does not function properly, its functions of predictive policing are non-existent, and its data integration capacity is fragile (Tribunal de Contas de Estado, 2016). Besides the contractual problem with Microsoft, the Court pointed out that the analytic systems were not integrated into the Detecta, some computers didn’t work properly, there weren’t enough people working in the data processing, many police departments didn’t have access to the system, and mainly, the camera system wasn’t duly spread over the city.

37The argument that has been mobilized by the authorities since then, is that the absence of resources and enough ability to deal with this system, forced the rearrange of Detecta as an extensive public-private assemblage, a collective of surveillance initiatives at the municipal and state’s level, where the private sector had an important role in the scattering of cameras, develop systems, apps, and administrate some information. However, the same report mentioned above shows that the commission was well aware from the very beginning that the private sector would play a vital role in making the system run, which points to a veiled intention of the government to force private transnational interests into public security, and involve it in liberal and market-based forces and arrangements.

38In what concerns its public dimension, Detecta came to represent an umbrella system for scattering cameras and data integration, producing statistics and heatmaps assisting police forces in the subsequent resolution of crimes – in short, situational awareness. On the other hand, it integrates a system of public cameras, from the Radar (linked to OCR systems), the municipal system of Cameras and images, the City Cameras, the CICC (Center of Integrated Command and Control), and a system of private cameras of residents which adhere to the system, through associations, and private companies. Moreover, the Detecta incorporates the SP+Segura system, which is an application dubbed "Waze of Security", in which citizens are instigated to denounce and report illegalities throughout the city, acquiring medals and virtual decorations, helping the authorities to map the illegalities in the city. In our analysis, we verified that both the dynamics of the expansion of the camera system is commanded by the private sector, which begins to assume a decisive role in this relationship with Detecta’s public dimension. However, this does not necessarily mean that Detecta is flawed, on the contrary, it configures a relationship in which public-private symbiosis is determinant in the security governance process, as we will present further.

39In this spirit, we understand that it is insufficient to solely analyze Detecta, ignoring the set of agents in its surroundings. Their relationships around Detecta are decisive to comprehend the effects of this system over São Paulo’s security, be them projects in the municipal, national or transnational companies, such as Microsoft, Genetec, Techvoz, Seg D'Boa, Tacira and Aster, their operators (Civil Police, Military, Secretariat of Public Safety), or users, such as associations and institutions. The relationship among these actors seem to structure a powerful and symbiotic network where the lines of public and private security, business and services, and even human and non-human agents are blurred, which demands from us not an approach over the essence of these parts, but in the set of relationships and strategies produced by this assemblage. It is a symbiotic relationship, once the state and municipality depend on the private digital and physical infrastructures, on its technologies and business practices of selling cameras. On the other hand, these companies depend on the legal authorization of the state, for public-private partnerships, and access to certain spaces and institutions. They become so entangled, that is difficult to distinguish what is a public service from what is a commercial practice, whether a technology by itself is capable of perform security or hidden sociotechnical practices are responsible for it. Then, we believe that the best set of tools that we could adopt to analyze it would be the idea of dispositif formulated by Michel Foucault. The idea of dispositif allows perceiving a set of relationships established between heterogeneous elements, which involve discourses, institutions, architectures, regulations whose purpose is to respond to an urgency. The dispositif would have a strategic function or remain in a perpetual strategic fulfillment process, which means to a large extent, that it produces effects that were not initially foreseen, but that sustains its activeness (Castro, 2016, p. 194).

40Several authors have been using different concepts to understand the compositions that come from the participation of multiple actors and processes in the management of security nowadays. Williams (2016), Holmqvist et al. (2015), and Sassen (2006), has been focusing on the concept of security assemblages, derived from the reified notion of agency in Deleuze and Guattari (1995). We say reified, because the concept of “agencement” was translated into English as "assemblage", which authors like Nail (2017) and Lemke (2018) point as limiting, since it merely denotes the reunion or union of things, being a concept closer to machines, defined only by their external relations of composition, mixing and aggregation. In this sense, the concept moves away from the notion of agencement, which describes a transitive notion of schematization, arrangement, or construction, associating the capacity to act with the gathering of things, a necessary condition for any action to occur. However, as Lemke (2018) explores, both notions of agency and assemblage would not be able to reproduce the complexity and breadth of the notion of dispositif, since it comes from Foucault's intention to understand how dispositif act on relations, in mechanisms and power games. Thus, the idea of ​​dispositif denotes three specific meanings, as pointed out by Foucault (1994, p. 299-300), first, demarcates the heterogeneous set of several elements (discursive and not discursive) in a network; secondly, it demarcates the nature of the relationship that can exist between these elements, that is, the game, the changes of position and the modifications of function that continuously operate in this dynamic; and finally, it attributes a historical character to this formation, since it understands that it seeks to respond to a certain urgency, being endowed with a strategic function. This way, we focus on the concept of dispositif rather than the concept of assemblage, not only for the reasons described in the body of the text, but for the relevance of the theoretical-conceptual question we present here, and for the potential analysis that it gives to our object, to attribute to this formation a strategic and dynamic character.

41The organization of São Paulo’s public security seem to work as a dispositif, composed by police institutions, private companies, public and private enforcers, surveillance technologies, security perimeters, citizens, all organized in a way to give space to a sort of strategies. Since the Dictatorial period in Brazil (1964-1985) the structural organization of the public security administration in São Paulo has not changed much, it still characterized by a strong hierarchy and division of activities between the police forces: while Military Police is directed to protect the civil society through patrols, acting over flagrant, the Civil Police is in charge of investigative activities. As Costa and Lima (2014) point out, there isn’t a proper concept of public security developed in the Brazilian legal order, and all the subjects related to it comes from the 1930s or were complemented in the 1960s during the dictatorial period. Costa and Ribeiro, understanding the public security as a field with proper characteristics and dynamics, affirm that it is very heterogeneous, being influenced by several institutions and organizations, but the military still one of its most influential organizations. Under this influence, the police adopted repressive practices – such as prisons, and massive repression against patrimonial crimes, and drug trafficking – as dominant paradigms to deal with criminality, in detriment of forms of prevention and control, as Sinhoreto et al. (2016) point out.

42The introduction of Detecta in this dispositif seem to scatter a model of surveillance, strongly concentrated in middle and upper-class neighborhoods, produced jointly by private (transnational) and public institutions. In this sense, Detecta seems not to change drastically the way security is provided, but actually to repose and reinforce the parameters and characteristics of this apparatus, whose set of "traditional" relationships is described by Marcos Alvarez (2004) as elitist, excluding and usually violent against marginalized social segments. This may be seen from the data published by the Security Secretariat of the State of São Paulo (2018), since the adoption of the Detecta system (2014), there has been only a slight fall in robbery rates in the city (from 209,536 robberies a year to 186,078, in 2017), in guilty homicides (from 587 to 423 in 2017). However, there was a large increase in drug trafficking (from 6521 to 9,173 in 2017), and a huge jump in the occurrence of police violence, not only in the city but in the whole state (which jumps from 369 occurrences in 2013 to 939 in 2017). Contrasting with the arguments of several interviewed, from the Policia Militar to the high levels of the public administration what we can notice is that the introduction of the Detecta may not be directed to assist the combat against the criminal and violent practices “preventively” in the city, but mainly to assist investigations, to monitor areas of illegal market, and with social disturbances.

Graphic 1: Crimes Committed in upper middle-class and rich neighborhoods in São Paulo, as % of residents

Graphic 1: Crimes Committed in upper middle-class and rich neighborhoods in São Paulo, as % of residents

Source: (Secretariat of Public Security of São Paulo, 2018)

43Even though some of the interviewed sustain that the Detecta has made police faster and even more efficient, it seems that it was insufficient to produce a substantial reduction in crimes in the city. Actually, the Graphic 1 shows that, in contrast with the argument of effectiveness of the Detecta deployed by the authorities, the crimes committed in some of the wealthiest neighborhoods of São Paulo – spaces that have adopted Detecta, City Cameras, Vizinhança Solidária, or initiated their adoption – such as common robberies, homicides, drug trafficking, have slightly raised even with the introduction of the Detecta in some of these spaces (from 2014 on). At the same time, the statistics of the Secretariat, from the last ten years, shows that the same crimes, mainly homicides and rapes, still substantially affecting poor neighborhoods in São Paulo, such as Jardim Herculano, Capão Redondo e São Mateus – places that were not considered to receive the Detecta – which corroborates with the arguments of Miraglia (2011) that violence still geographically distributed in an unequal way in the city. This is not to affirm that Detecta is responsible for the increase of the crimes, rather, to say that its expected deterrent effects are not yet apparent, and its effectiveness may be questionable. One can argue that crimes seem to raise, once the system was able to register it more accurately, but it would be inaccurate once neither all neighborhoods adopted the Detecta system’s at the same time, but they register an increase in criminality at the same time, and moreover, neither all crimes occurred were recorded by cameras, but denounced afterwards and processed in the system by the officers.

  • 3 About this, one of our interviewed, a coronel of the São Paulo’s Military Police responsible for th (...)
  • 4 According to Holston (2013), the “insurgent citizenship” is a perception that questions the regimes (...)

44In fact, most of the discourses recently deployed about the effectiveness3 – often concentrated in the improvement of the agility of the agents to respond calls, to produce “situational awareness”, and to solve occurrences – of this surveillance system were not related exclusively to the high rates of crime, robbery, and homicides in the city, but to its capacity do deal with problems that allegedly would arise from what Holston (2013) call as “insurgent citizenship” in the city, often manifested in the multiple and intense circulation of lower income people and groups of diverse origins and social conditions in spaces previously exclusive, such as shopping malls, banks, airports, etc, and even in wealth neighborhoods, and parks4 – this places were not “reserved” for high income groups, however, due to a previous extreme economic inequality, and to an urbanistic model that historically expelled the poor to live in peripherical regions of the city, these are areas not frequented intensively by them. In view of this, the Detecta appears as an instrument for governing and modulate the circulating flows of the city, allowing the authorities and private companies monitor “social disturbances”, and patterns of circulation in certain perimeters, potentially acting on problems like: the political instability and the political risks that allegedly would arise from popular manifestations; the agglomeration of people and risks of terrorist attacks during major events that would be based in the city; the confrontation with what the government labels as processes of great social degradation, as in the case of the region called “cracolândia”, where there is a crescent struggling between real estate interests – that seek to economically explore that space – and the permanence of homeless and drug-addicted people in the region of Luz.

45Then, with crime rates increasing in several neighborhoods in São Paulo, even after the introduction of Detecta, and with its spread over wealth and upper-middle class neighborhoods, spaces with high dense circulation, and in spaces with severe social conflict, the Detecta, instead of been related to the prevention of crimes, seems to reinforce the organization of the security as a response to the circulation of unwanted groups in the city – a demand from wealth neighborhoods associations, private companies, local business – ceasing or modulating it in certain public spaces. Then, in our view, the diagnosis promoted by the authorities to sustain the implementation of Detecta is not directly related to the reduction of the crime rates, but actually, intends to respond to the intense circulation and mobilization that would allegedly produce it in some spaces. Therefore, enforcing the capacity of governing people in public spaces, and consequently been able to interrupt its circulation, producing aesthetic spaces favorable for the producing of security, seem to be the strategic fulfillment that permanently orients this dispositif.

  • 5 About this, the security and monitoring company Aster developed a study called “Janitoring together (...)

46In view of this, Detecta contributes to the formation of what seems to be a security dispositif, which consists of a set of public-private conjunctions, information and communication technologies and infrastructures, that together with an unprecedented civil participation, operates a discourse of fear on circulating flows of people, at the same time that introduces perimeters of intense surveillance and control. This discourse is pronounced countless times by authorities, businessmen and managers, in order to securitize circulatory flows of people in the city5, understanding that any unregistered form of circulation, any deviant form of circulation of people in determinate neighborhoods – together with the police inefficiency, the absence of civil society participation in the process of vigilance and denunciation – may be considered a menace and a trigger to the intensification of illegalities and disorder. For the internal stabilization and expansion of this dispositif, daily violence needs to be constantly addressed and recalled, the state of mistrust that marks the policing strategies in the culture of control must be permanently maintained. This perspective on the dispositif matches with the idea of securitization of social issues explored by Bigo (1995), in which the increase of the capacity to govern security depends on the production of insecurity in the society itself. In other words: "this insecurity will be translated into a social demand for an intervention of coercive state agencies (...) the processes of securitization and insecuritization are inseparable" (C.A.S.E. Collective, 2006, p. 19). Strictly, also governmentality in Foucault, as explored by Lemke (2017) and Optiz (2011), presupposes a government (through) of fear, that is, the constant production and reproduction of threats to freedom, as a means of expanding the technologies available to combat and "manage" such threats.

47As mentioned by Bauman (20013 p. 123), and commented by Lyon (Bauman, 2013, p. 96), security, its practices and discourses focus on “external objects, visible and registrable” trough the surveillance apparatus, which tend to be “blind to the individual reasons and choices underlying the recorded images; therefore, they might lead to the substitution of the idea of evil individuals for “suspected categories”. This points to the argument that visibility and register are central aspects for security practices, meaning that any invisible spaces (black spots), uncapable of registering movements, may threat the stability of the surveillance apparatus, and of the exercise of security. This binds the dynamic of permanent expansion of the Detecta system to the effectiveness of security, while shifting the focus of security from distinct objects, to cloudy and shapeless risks.

  • 6 The term janitorial is an immediate translation from the term “zeladoria” in Portuguese, which enco (...)

48As a response, this dispositif proposes the scattering and integration of public-private cameras which, on the one hand, would help to reduce the costs of the management of public security through an automatic and constant monitoring of the city, and on the other hand, involves the citizens on an active exercise of vigilantism, stabilizing the dispositif while meeting their local demands. Projects such as Genetec's Citiwise and Microsoft's Public Safety & National Security, present in the districts of São Paulo, sustain that the security government would only be possible if distributed between security agents and residents, once the latter is responsible for surveillance and janitorial6 of their territories through the cameras of security. The expectancy is that citizens get involved in security issues of their communities, buying and installing cameras, denouncing illegalities committed both by residents or “outsiders”, and this could be made by uploading images, videos and other material through Whatsapp groups, or using the app SP+Segura, developed by the municipal administration. In a report produced by the SAAP (Society of Alto de Pinheiro’s Friends), a neighborhood association from one of the wealthiest neighborhoods of São Paulo points that during a meeting, with police and security entrepreneurs, pointed:

Are you scared of walking in the streets? Have you witnessed a violence situation? Do you know how to act when spotting a stranger in a suspect attitude? Thus, together with the public administration, the citizen also exerts a fundamental role in the maintenance of the neighborhood where he lives. (…) The captain Cunha Neto highlighted the effectiveness of the residents’ WhatsApp groups. The tool allows them to inform each other about suspects activities in your streets (Saap, 2016, p. 02).

49This excerpt demonstrates how population is instigated to participate as an active agent in the production of a “safer” space, acquiring cameras, and becoming vigilants, otherwhise, it may result in some consequences, these publications always tend to point. About this, an article of the magazine Veja São Paulo, discussing about the deployment of smart cameras in some neighborhoods of São Paulo, published:

The Jardim Lusitânia, nearby the Ibirapuera Park, also belongs to the list of regions with a “Big Brother” scheme. (…) On May 11th, three of these cameras in Macau street spotted an action of thugs in a house. (…) In an investigation about the crime, the images were demanded by the police, and it show three man arriving by car and invading the residence. “The robbery occurred exactly in a spot without monitoring. That’s why, the images are from far away. We couldn’t identify the plate of the vehicle used in the action, neither the faces of the criminals”, states the deputy (Soares, 2017)

50As stated by Amicelle et al. (2015, p. 46), security devices perform security from the reconfiguration of social spaces, redefining borders and redistributing meanings in networks of relationships. These neighborhoods have been reorganizing the way they deal with security, using instruments like Camera images, WhatsApp and other apps to report what they consider to be illegal practices, and getting close to police officers, at a point that they have access to their private numbers. At the same time, private companies seem to intermediate these relationships, constantly promoting events on technology and public security, both directed to businessman or to the public in general.

  • 7 As an example, some of the local companies that provides surveillance and algorithmic based cameras (...)

51Among the events in public security in 2018 and 2019 that sought to present Detecta and other urban surveillance projects, four of them stand out: The International Security Conference & Exposition (ISC); the Latin America Defense Security (LAAD Security); the “DroneShow” and, finally, the Condominium Security Symposium organized by the state congressman "Coronel Camilo". There, the "networks of experts", as explored by Bigo (2008), mostly security and electronic security businessmen, defense companies and security agents, are in charge of the maintenance of a "semantic continuum" regarding the way security is managed, operating discourses of fear and exploring possible solutions. In these events, the discourses that mobilize the network of actors and the expansion of the dispositif are performed in lectures, conversations, institutional materials and specialized magazines. Based on Bigo's (2008) approach, these events demonstrate how (in)security becomes an activity managed by “experts”, organized in global chains of defense and security7 that:

(...) claim, by the “authority of statistics”, to have the ability to rank and prioritize threats and determine what constitutes security exactly. (...) Security is thus conceptually reduced to surveillance technologies, information extraction, coercive actions against social and state vulnerabilities, in general, a form of general survival against threats from different sectors, but also, security is disconnected from human, legal and social guarantees and individual protections (Bigo, 2008, p. 12).

52Thus, the focus of their speeches and diagnosis is never directed towards the understanding of the threats with relative depth, but always pointed to the certainty that the gains of efficiency in police operations, and a regime of visibility will be translated into the reduction of the discomfort. In the ISC Brasil 2018 event, it was common to find specialized magazines stimulating fear, the continuum of security (as if terrorist threats were crime-related in Brazil) sustaining conservative agendas in order to deal with the security problems. As presented in Figure 1, some of the headlines addressed: “Infrastructure: how a structural collapse can impact corporate and personal security planning”; “Terrorism: special edition on the attacks on the French newspaper and parallels with the Brazilian reality”; “Domestic terrorism and private security”; “Managing Uncertainty: The Art of Risk and Security Manager”; “Age of criminal responsibility, an enigmatic trilogy: maintain, reduce or extinguish ...”. In general, the headlines sought to merge, moral panics, discourses of fear, “taboos”, and suggest notions of entrepreneurship and “technophilic” solutions to deal with it.

Figure 1: Magazines Distributed in Electronic Security Fairs

Figure 1: Magazines Distributed in Electronic Security Fairs

Source: (Author’s collection, 2018, São Paulo)

53As a smaller version, a condominium security event brought together in one room security agents, representatives of Community Safety Councils (Consegs), businessmen, politicians, maneuvering the same security-surveillance logic, but emphasizing the participation of the community in the division of costs and responsibilities for governing security. During the lecture of a Military Police commander about the series of crimes that occur around that business condominium, the audience (residents, trustees, presidents of associations and merchants) was asked to reflect on the importance of acquiring surveillance and security equipment and encouraged to join the Vizinhança Solidária program (Solidary Neighborhood). This one, a project headed by the state deputy “Coronel Camilo”, which organizes civilian vigilantism in their neighborhoods, helps to create an environment of engagement in security among them, as well as for the introduction of electronic surveillance systems. The brochure of this program presents for the residents recommendations that ranges from “don’t be indifferent to what happens around you”; “Be friendly [with employees], but discreet”; “Check the signs of danger in your neighborhood”; to "attention to common disguises”, “be a good observer”, “participate in Consegs” and especially: “install security cameras”, because "”the cost of this equipment can easily be prorated by the residents of your street or condominium. This visual control is fundamental since an image can be decisive to prevent, repress or investigate a crime” (Camilo, 2018).

54It is precisely in this aspect, the sociotechnical organization of communities in middle and upper-class districts, that a new roll security alarms and camera system companies begin to establish themselves. In general, they will mediate the relationship between residents’ associations, technology companies, the state and municipality, on the one hand, guiding the acquisition, provision, installation of cameras, as well as communication systems among residents, in some cases even training the residents to identify problems, on the other hand, ensuring that the images of these cameras can be accessed and used by the state and municipality – also selling the data analytics acquired to the state. One of the biggest companies in this field understands its activity as a “social project” since it gives the sense of empowerment among the residents to deal with problems related to public safety. In an interview with an executive of this company, this idea of social empowerment, rapprochement between residents, and the security dispositive is clarified:

It is the social bias that guides [us]. For our biggest problem, from the beginning, was not having any kind of political interference, the political may be changed, but not the social policy. Then, we use everything that belongs to the community, and they understand that they are an active agent now, and more, one offers its camera to have access to all the others, so the difference is that it leaves the self [action] to become a collective one, with a multiplier gain (Executive C1, 2018).

55The introduction of Detecta modifies a series of relationships, committing and empowering the private sectors and associations of residents to assist in “combating crime”, activating them as a part of the security dispositif. But how this impact on the landscape of the city? How empowered citizens, security agents, and algorithm-based systems combine themselves in an assemblage that intensifies the process of segregation in the city?

Newly Public-Private Security Perimeters, old Segregational Patterns

56Considering that Detecta system did not integrate the whole functionalities of a predictive surveillance device, with analytical readers and automatic alerts, what is left is an apparatus with ample capacity of surveillance and governing citizen conducts, but dependent on a human eye for its completeness. Automation is nowhere ‘complete’, neither in São Paulo nor in New York, always depending upon some kind of human agency that unsettles the pipe dream of autonomous and technological perfection. In São Paulo, these systems are strongly dependent on public, private and civil security agents in monitoring processes, and producing information for the government strategies. This surveillance process is aided by the use of mobile phone applications made available to residents in security perimeters, which is not only directed to the monitoring, but to spot dangerous areas, to suggest better paths for the users to take, to have an exclusive channel to communicate with the private security agents or the police.

57The subjective look of the security agent or of the vigilante citizen completes the surveillance apparatus with his “intuition” or “experience”. Through this process, an extremely sophisticated surveillance system meets the tradition of prejudice and discrimination that marks the São Paulo Military Police activity, as explored by Alvarez et al. (2004) and Caldeira (2016). The authors point to the fact that even after the democratization in São Paulo, the criminal justice apparatus has a strong racial bias, the security policies regarding the public security remain elitist (strongly influenced by local elites), which leads to the emergence of violent police practices who customarily targets black and poor people, and to the formation of patterns of segregation in the city, whose most significant one was described by Caldeira (2016) as the formation of fortified enclaves, private condominiums apart and inside the city under the watch of private security companies. São Paulo police’s racist postures are also perceived by the numbers, as a study from the Forum Brasileiro de Segurança Publica shows, that in 2013, 56.3% of those killed by the police were black, a number that raised to 65% in 2015, and decreased to 64% in 2018 (Martins, 2019). The Detecta, therefore, is ultimately manifested as a sociotechnical assemblage in which the human gaze and its subjectivities, will be incorporated into an amplified and sophisticated visualization system, reconditioning it as an apparatus to modulate and constraint unwanted patterns of circulations in some spaces. This happens once the system depends on the gaze of the vigilant, because it didn’t incorporate the image analytics in all of its cameras, then not only private guards (as we will demonstrate further) will complete this system with your gaze, but also residents in neighborhoods are stimulated to be vigilant, as we have discussed above. Indeed, the culture of control inscribed into Detecta helps to manage various aspects of the society as a means of inhibiting crime, employing surveillance technologies as a way of amplifying this government over people and things. However, as Garland (2008, p. 288) realizes, this culture “lives well with economic and social policies that exclude entire population contingents, provided such segregation makes the system operate more harmoniously”.

58It is precisely this effect that we observe in Detecta, an apparatus that fits perfectly with the security dispositif by combining the customary discriminatory practices with an extended and distributed regime of visibility, controlling and modulating the circulations in specific environments of the city. This happens due to a dynamic of expansion of the camera system governed by the private initiative, which focuses on middle and upper-class neighborhoods in São Paulo, business centers, strategic areas, and in some public or private spaces of high public circulation, forming what we call “security perimeters”. In this matter, a local entrepreneur, who provides camera systems and “training” for the neighborhoods told us that “who maintain it are the residents, who pay a monthly tax for its installation, and to have access to these images, and in contrast, they provide us internet, the images and the energy for the cameras stay on line permanently” (Executive C1, 2018). These perimeters are formed through the agency of national and transnational electronic monitoring companies, who convoke meetings with residents' associations, merchants and security agents, where they offer electronic solutions as effective systems to deal with the threats and its diffusion. They provide camera installation services, the provision of mobile applications for image access, communication between residents, sometimes training and exclusive access to private security guards or to the police. As one of the highest public administrators told us about the System City Cameras – which we understand as one of the instruments that composes Detecta –

the companies and the city or the state celebrate an agreement after responding to a bid notice: (…) We are with 25 companies already registered, which brought the documents, all set. (…) The Cameras with LPR/OCR are extremely expensive (…) then what is our idea, is that the people who live in community, they could also install cameras like this, and we begin together working this issue (Secretaria 1, 2018).

59Whenever all these cameras are installed, the residents interconnected, and a bit familiarized with the practices of the Vizinhança Solidária, we understand that this spaces – that could be a street, a neighborhood, several blocks, a temporary area for street carnival, the contours of a shopping mall or a huge hospital – become a security perimeter, mutually governed by both public and private agents, with some space by janitorial actions by the residents or local businessman. In these perimeters, the fear from crime, and from the circulation of unwanted people is constantly stimulated by electronic companies, police officers, and by the informal communication in chat applications. This process together with the empowerment of the residents or workers to visualize, identify and “act” over certain “threats”, helps to produce some adverse effects that will be further highlighted. As Melgaço (2010, p. 105) points out, often the sense of insecurity mobilized in certain spaces is disproportionate to real risks (see the reduction of homicide rates in São Paulo, and the maintenance of discourses on “chaos and barbarism”), and this forms a “psycho-sphere of fear” where ideas, beliefs, passions are determinant in the production of meaning for the inhabitants. According to the author: “the psycho-sphere of fear thus appears as a precondition and justification for the installation of a security technosphere. This technosphere concerns all forms of technical materiality around the security ideal and obviously includes securitization processes”.

60Fueled by the availability of private surveillance resources for the expansion of the security dispositif, this psychosphere of fear potentiates tensions related to the circulation of “undesirables” in these perimeters. As already mentioned in the “Vigilância Solidaria” primer, we notice that the circulation of people, as well as the multiplicity of adverse relationships that would lead to crime, are the objects to be securitized and maintained in permanent surveillance. The images below (Figure 2) represent these perimeters, respectively, the first one is a shoot from a private security monitoring company’s platform, it represents one perimeter in a neighborhood, where the blue dots are cameras, with their range. Whenever these private security agents desire, they may access the cameras by clicking on the dots. The disposal of these cameras makes visible a strategy of building widely surveilled security perimeters, where every entrance of a determinate space is monitored by cameras similar to the one in the second picture, shot in one of these perimeters in Jardins neighborhood in São Paulo.

Figure 2: Examples of public-private Security perimeters

Figure 2: Examples of public-private Security perimeters

Security perimeters sign

Security perimeters sign

Source: (Author´s collection, 2018, São Paulo)

61As Lyon (2018) points it in his discussion of the new culture of vigilance that organizes contemporary capitalism, social and urban dynamics, the individual becomes an extremely relevant node in the network, since he deliberately produces information revealing a series of data considered precious for data analysis companies. In the same way, the inhabitant of these perimeters becomes a fundamental component of this security dispositif, not only as receptacle-amplifier of the discourse of fear that makes it grows, but as a producer-consumer of this surveillance system, a sort of essential hub that produces information through the disposition and sharing of the links of his cameras – extending his look on what is considered adverse – and permanently consumes the information produced by his and the neighbors’ cameras. His untrained eye, conditioned by the perspectives of permanent suspicion (Larssson, 2016) of community WhatsApp’s groups, and private consultant’s lectures, makes prevail discriminatory practices against “undesirables” in these perimeters.

62In some cases, like in Vila Madalena during the street carnival, one of these perimeters equipped with “analytics” was tested, translating the residents' specific fears into algorithms of image readings. This can be glimpsed from the speech of an executive of a company that provides this type of services:

So, when we had these images live, in real-time, with operation inside the container, we caught drug trafficking, minors using alcohol, vandalism, several actions that are inherent in any society. Because when we have a lot of people, they think they are immune, protected by the collective, and that is our great challenge, studying behavior and getting the camera and technology to make that kind of assessment. Today we can manage with some tranquility, car volume, people, and when the camera shows above (Speed Dome) in Largo da Batata, the analytic counts in seconds, how many people are per square meter. (...) This year the issue of bringing people to the quadrilateral was successful, without running over the limit number, and the bars won with it, the trade won with it, the community that enjoyed the carnival, won with it, and those who went out to vandalize, arrived late and they did not enter. Whoever was not wanted was left out. And it was good for everyone (Executive C1, 2018).

  • 8 Regarding these “fears” and “demands”, an article recently published by Forum Magazine (2019) revea (...)

63In another perimeter of São Paulo similar systems are adopted, with image analytic system from private companies. This system, an algorithm of image analysis, accuses the invasion of a certain space, or the permanence of an individual there by a certain time, issuing an event alert to private agents, as we can see in Figure 3 below. This specific analytic is known in the market as “Loitering”, whose most direct translation into Portuguese would be “vadiagem”, a term that in Brazil refers to a 1941 law considered the "legal inheritance of dictatorships", and an elitist instrument for the subjection of the working classes commonly known as the “Lei da Vadiagem”, which criminalizes the idleness of the individual, who in full force for work is circulating in public spaces. During the dictatorial period, this law, which is still in force in the country, was responsible for several arrests in flagrante delicto in metropolitan areas, especially when the individuals were undocumented, which included the work card (until then a symbol of citizenship) (Villela, 2014). It is not surprising that this analytic seems to re-enact this subjugation practice in the form of a sequence of “source codes” that will compose the algorithm, justifying preemptive suspicions about people circulating in these securitization perimeters, attending to the demands of residents backed by the psychosphere of fear of prejudice8.

64The algorithm is a product of human agency and action, the result of a series of interactions, disputes between values, interests, and programming that, as Latour (1991) asserts, are crystallized in artifacts. Depending on the arrangements between actants, that range from data scientists, companies, security agents, residents, camera systems, algorithms and the past data that “trains the algorithm”, a particular configuration of the analytics are developed and end up embodying several clientelist views of security. In this sense, these analytics impose a certain pattern of conduct considered as normal by the developers, and anything that does not fit on that could be subjected to suspicious alerts.

Figure 3: “Loitering” Analytic in São Paulo

Figure 3: “Loitering” Analytic in São Paulo

Source: Extracted from the TV show “Mais Você”, 2017, São Paulo

65A case in which these assemblages between biased views and a system of increased visibility by the cameras is clear, is that of one public university of São Paulo, a security perimeter formed during the expansion of Detecta in early 2018, composed by a multiplicity of applications and cameras distributed across campus, and digital mediation platforms, allowing the territory to be electronically controlled from a monitoring center. With all of its accesses watched with cameras that can incorporate analytics, the university provides students and servers with a mobile application that includes a "panic button" that can issue alerts in the event of an incident. The alert issued is reported on a Google platform map on a screen, and the perimeter cameras are triggered. Sometimes, as the head of the control room told us, the Military Police send them an alert for a specific vehicle, or wanted individuals, and they cooperate with the persecution. However, such a procedure is so rare that it often requires the head of security to demonstrate its working to visitors. In general, the routine is the monitoring of the cameras by agents, who, as they say, have the “experience” to identify suspicious deviations and behaviors. As one of the of security crew, while narrating an episode, highlighted to us:

  • 9 The real name was changed, once it is related to a third part.

There’s a training to perceive what is a suspect attitude. André9 who is our employee of the month, he is busting up everybody. Just like we’re doing it here, monitoring this perimeter, he noticed a suspicious attitude from a guy who had been here somedays ago. (…) Monitoring the cameras, he noticed when the guy comes close to the bike rack (…) he came close, and usually it is simple to characterize a student from another individual, and then, when he arrived close to the bike, André had this conception [insight], and started to follow him, when he took the bag and stole the bike. (Security S1, 2018)

Figure 4: Videowall of the University`s Control Room

Figure 4: Videowall of the University`s Control Room

Source: (Author’s collection, 2018, São Paulo)

66The cameras, however, were always pointed at the same people, young black or brown people who “diverged” from traditional student stereotype, as we can see in Figure 4. The image in the figure above is a record maintained by the security from a group that they consider suspect, as the head of the security told us: “These are images that we record from the field, there are some very young boys that sometimes we don’t even care about them, but we tell that they approach people to steal them, it’s unbelievable” (Security S1, 2018). Even though these cameras were previously fixed in “strategic” points across the campus by private security agents, and usually the images are in automatic loop, the private security agents in the control rooms usually assume the control of the cameras, focusing, zooming in and out, select specific cameras to spot any detail they may consider relevant. The cases demonstrated also involved robberies committed by people with the same characteristics, making a recurring pattern appear.

67This exposes the fact that cameras arranged in strategic places on campus, besides providing support for investigations, often tend to point to a type of unwanted circulation, aligned with an aesthetic pattern of the peripheries, reinforcing a legitimized by a regime of visualization in its records. On the other hand, what is interesting is that this regime of visibility exposes behaviors, deviations and practices that do not necessarily constitute a crime, but also do not fit into the acceptable patterns of this culture of control, which makes them capable of identification and modulation by private agents in the name of security, and a more “efficient” campus.

68These cases seem to illustrate how the inscription of a culture of control in a surveillance apparatus like Detecta – understood as a public private assemblage, a collective of initiatives in several levels – produces a state of permanent distrust and creates perimeters of exclusive security. On one side, the prejudice of residents of the security perimeters seems to be composed in the cameras with analytic algorithms, making suspicious governed by algorithms and private agents permanent and durable. On the other side, in the perimeters without analytics, the state of suspicious is perpetuated by the assemblage between biased eyes and wide camera systems. These cases evidence how customary discriminative practices are reproduced in a context of hypervigilance, and algorithmic mediation, once this system allows private companies, and biased residents to govern the circulation of people in certain areas, interrupting it whenever necessary. Thus, in an attempt to deploy sophisticated forms of control, situational prevention and awareness, the São Paulo’s government and all the assemblages that constitute the Detecta, ends up by making durable and permanent the discriminatory practices already present in the city.


69This research sought to explore how the inscription of police and military concepts in surveillance systems enables new ways to govern the city and perform security. Moreover, we intended to explore how new information technologies have conditioned security as a dispositif of control and illiberal governmentality, producing fears, establishing patterns of normality and diversion, and acting upon them. The sociotechnical and public-private assemblages that makeup São Paulo’s security dispositif end up establishing an illiberal form of governmentality, where the persistent intervention and arbitrariness of the state and other agents are justified in terms of maintenance of freedom (Optiz, 2011).

70Both in New York and in São Paulo, the spread of a culture of control through the concept of smart and safe cities helped to equate the idea of security to safety, quality of life, and mobility, introducing apparatuses of hypervigilance. It is in this sense that the exercise of security is no longer a way to impose disciplinary control, but figures as a form of governing the deviations, in which all are monitored but only what is considered threatening "freedom" (for some people) is subject to redirection or modulation. This government is distributed among public and private security agents, algorithms and sociotechnical apparatuses, residents and workers in security perimeters, endowing them with a capacity to stipulate “normal" behavior and pervasive deviances (Rouvroy and Berns, 2018).

71It seems that with all technology brought up by Detecta, trying to reorder policing as a data-based activity, the only visible result of its introduction is the active role of national and transnational companies – such as Microsoft and Genetec, and the private national companies connected to Global Securalliance, which are helping to create a culture of surveillance in São Paulo while providing crucial systems to the functioning of Detecta – in the modulation of the public policing and surveillance. This wide role of the private sector in the management of public security reduced the burdens for the state and municipal administrations, opening another space for the private accumulation in the security sector. At the same time, it allowed the adaptation of Detecta in a more suitable manner to the security dispositif, since by no means it eliminates arbitrary and discriminatory practices from that, but incorporates it through socio-technical assemblages that maintain active and circulating the elitist's interests of portions of the population in the city. Moreover, as pointed by David Lyon (Bauman, 2013, p. 96), the security of today generates as a sub product – or as a deliberated policy – certain forms of insecurity usually felt by poor people that these same measures should protect.

72The formation of security perimeters is something that was not unheard of in the city, as Caldeira (2016) explores in her writings about fortified enclaves (condominiums) formed between 1980 and 1990. At that time, these walled enclaves were directly oriented to block the circulation of people, with a strong-armed presence of private agents, promoting a kind of insularity from the “outside’s life”. These contemporary security perimeters, in contrast with the enclaves analyzed by Caldeira (2016), do not block or avoid the circulation of people in the perimeters, rather it enables a modulated circulation, while requires a vigilant and janitorial attitude of the residents which extrapolates security issues, governing multiple actions of the neighbors. This way, the appearance of security and control, the exposition of cameras and warnings are understood as a performance to modulate the circulation of people in these spaces, diffusing an image of an empowered and vigilant community.

73Then, the introduction of the Detecta system in São Paulo, as an effect of surveillance, reconciles private, clientelist and elitist interests into a public security dispositif, stimulating an appearance of security, while normalizes indirect forms of violence such as old prejudices, mistrust and segregating practices. The vigilantism and the formation of security perimeters appear to be disproportioned responses from the Detecta to neighborhoods and spaces that have relatively low crime rates, yet it empowers certain social groups and private companies to govern security, making it even more porous and permeable.


75The authors would like to thank FAPESP (Fundação de Amparo a Pesquisa do Estado de São Paulo), for all its support to this research.

Haut de page


Adorno S. et al., 2008, « Homicídios, desenvolvimento socioeconômico e violência policial no Município de São Paulo, Brasil », Revista Panamericana de Salud Publica, nº 4, p. 268-276.

Alvarez M.C., Salla F. e Souza L. A. F., 2004, « Políticas de Segurança Pública em São Paulo: uma perspectiva histórica », Justiça & História, nº 8, p. 173-199.

Amicelle A., Aradau C. and Jeandesboz J., 2015, « Questioning security devices? Performativity, resistance, politics », Security Dialogue, nº 46, p. 293-306.

Amoore L., 2013, The Politics of Possibility: Risk and Security Beyond Probability, Durham, ed. Duke University Press.

Aradau C., 2010, « Security That Matters: Critical Infrastructure and Objects of Protection », Security Dialogue, nº 41, p. 491-514.

Arthur K., 2015, « Supporting Law enforcement resources with predictive policing », in Microsoft Government. [].

Assessoria Especial para Assuntos Internacionais, 2013, Missão a Nova York, Londres e Amsterdã/Haia: Segurança Pública, dir. Governo do Estado de São Paulo.

Aster, 2017, Zelando Juntos: Vila Leopoldina, São Paulo, Without Editor.

Bauman Z., 2013, Vigilância Líquida, Rio de Janeiro, ed. Zahar.

Belli B., 2004, Tolerância Zero e democracia no Brasil: visões da segurança pública na década de 90, São Paulo, ed. Perspectiva.

Bhandari P., 2016, « Predictive Policing: The future of law enforcement », Microsoft State and Local Government. [].

Bigo D., 1995, « Grands débats dans un petit monde. Les débats en relations internationales et leur lien avec le monde de la sécurité », Cultures & Conflits, nº 20, p. 7-48.

Bigo D, 2000, « Identifier, catégoriser et contrôler : police et logique pro-actives », in Bonelli L. et Sainati G. (dir.), La machine à punir : pratiques et discours sécuritaires, Paris, éd. L’Esprit Frappeur, p. 53-85.

Bigo D., 2008, « Globalized (in)security: The field and the ban-opticon », in Bigo D. and Tsoukala A. (eds.), Terror, Insecurity and Liberty: Illiberal Practices of Liberal Regimes after 9/11, London/New York, ed. Routledge, p. 10-48.

Borne M., Johnson H. and Lisle D., 2015, « Laboratorazing the border: The production, translation and anticipation of security technologies », Security Dialogue, nº 4, p. 307-325.

Brantingham P.L. and Brantingham P. J., 1993, « Environment, routine and situation: toward a pattern theory of crime », in Clarke R.V and Felso M. (eds.), Routine Activity and Rational Choice, New Brunswick, ed. Transaction Press, p. 259-294.

Caldeira T., 2016, Cidade de Muros: Crime, Segregação e Cidadania em São Paulo, São Paulo, éd. Editora 34.

Camilo C., 2018, Cartilha da Vizinhança Solidária. [].

C.A.S.E. Collective, 2006, « Critical Approaches to Security in Europe: A Networked Manifesto », Security Dialogue, nº 37, p. 443-487.

Cardia N., Adorno S. e Poleto F., 2003, « Homicídio e violação de direitos humanos em São Paulo », Estudos Avançados, nº 47, p. 43-73.

Castagnino F., 2018, « Critique des Surveillances Studies. Éléments pour une sociologie de la surveillance », Déviance & Société, nº 42, p. 9-40.

Castro E., 2016, Vocabulário de Foucault: Um percurso pelos seus temas, conceitos e autores, Belo Horizonte, ed. Editora Autêntica.

Chronopoulos T., 2017, « The Making of the Ordely City: New York since the 1980s », Journal of Urban History, nº 32, p. 1-32.

Clarke R., 1980, « Situational crime prevention: Theory and practice », British Journal of Criminology, nº 20, p. 136-47.

Côte-Boucher K., Infantino F. and Salter M. B., 2014, « Border security as practice: An agenda for research », Security Dialogue, nº 45, p. 195-208.

Deleuze, G. e Guattari, F., 1995, Mil Platôs: Capitalismo e Esquizofrenia, vol. 1, Rio de Janeiro, Editora 34.

Feldman, A., 2004, « Securocratic wars of public safety  », Interventions, nº 3, p. 330-350.

Felson M., 2002, Crime and Everyday life, Thousand Oaks, Ed. Pine Forge Press.

Forum, 2019 « Moradores de Higienópolis querem “limpeza” de população de rua », Revista Forum. [].

Foucault M., 1994, Dits et Écrits 1954-1988: Tome 3, 1976-1979, Paris, éd. Gallimard.

Foucault M., 2008, Segurança, Território, População: Curso dado no Collège de France (1977-1978), São Paulo, ed. Martins Fontes.

Garland D., 2008, A Cultura do Controle: Crime e ordem social na sociedade contemporânea. Rio de Janeiro, ed. Editora Revan.

Genetec, 2019, Citiwise. Disponible en, (], (consulté le 11/04/2019).

Geraldo 45., 2014, « Detecta: Tecnologia Contra o Crime », Youtube. [].

Governo do Estado de São Paulo, 2015, « Secretaria lança cinturão eletrônico de monitoramento do Detecta em todo o litoral de SP » [].

Graham S., 2016, Cidades sitiadas: o novo urbanismo militar, São Paulo, ed, Boitempo.

Haggerty K. D. and Ericson R. V., 1999, « The Militarization of Policing in the Information Age », Journal of Political and Military Sociology, vol. 27, nº. 2, p.. 233-255.

Harcourt B., 2007, Against Prediction: Profiling, Policing, and Punishing in an Actuarial Age, Chicago, Chicago Press University.

Holston J., 2013, Cidadania Insurgente: Disjunções da Democracia e da Moderninade no Brasil, São Paulo, ed. Compania das Letras.

Holmqvist C., Bachmann J. and Bell C., 2015, « Assemblages of war: police: An introduction », in Holmqvist C., Bachmann J. and Bell C. (eds), War, Police and Assemblages of Intervention, London, ed. Routledge.

Holston J., 2013, Cidadania Insurgente: Disjunções da Democracia e da Moderninade no Brasil, São Paulo, ed. Compania das Letras.

Larsson S., 2016, « A first line of defence? Vigilant surveillance, participatory policing, and the reporting of “suspicious” activity », Surveillance & Society, nº 15, p 94-107.

Latour B., 1991, « Technology is society made durable », in Law J. (ed.), A Sociology of Monsters: Essays on Power, Technology and Domination, London, ed. Routledge.

Latour B., 1994, « On Technical Mediation: Philosophy, Sociology, Genealogy », Common Knowledge, nº 3, p. 29-64.

Lemke T., 2017, Foucault, Governamentalidade e Crítica, São Paulo, ed. Politéia.

Lemke T., 2018, « Rearticulando o Conceito de dispositivo: Combinando STS e Analítica do Governo », Mediações, nº 1, p. 32-62.

Lyon D., 2018, « Cultura da Vigilância: envolvimento, exposição e ética na modernidade digital », in Bruno, F., Cardoso B., Kanashiro M., Guilhon L e Melgaço L. (dir.), Tecnopolíticas da Vigilância: Perspectivas da Margem, São Paulo, ed. Boitempo.

Martins L., 2019, « Em SP, 54% das pessoas Mortas pela PM no Ano Passado eram Pretas ou Pardas », in Uol [].

Melgaço L., 2010, Securização urbana: da psicoesfera do medo à tecnoesfera da segurança, São Paulo, Universidade de São Paulo,.

Mello M. M., W/D, PMESP: Planejamento Estratégico - Do Sistema Integrado de Gestão ao Plano de Policiamento Inteligente (PPI). [


Mesquita Neto P., 1999, « Violência policial no Brasil: abordagens teóricas e práticas de controle », in Panoli, D. (ed.), Cidadania, Justiça e Violência, Rio de Janeiro, ed. Fundação Getúlio Vargas, p. 130-148.

Miraglia P., 2011, « Homicídios: Guias para a interpretação da violência na cidade », in Kowarick P. e Marques, E. (ed.), São Paulo: Novos Percursos e Atores, São Paulo, éd, Editora 34, p. 334-339.

Nail T., 2017, « What is an Assemblage? », SubStance, nº 46, p. 21-37.

New York Police Department (NYPD), 2009, « DAS: Public Security Privacy Guidelines », [].

New York Police Department. W/D, « Improving Intelligence and Counterterrorism Capabilities », [].

Optiz S., 2011, « Governo não Ilimitado: O Dispositivo de Segurança da Governamentalidade Não Liberal », Ecopolítica, nº 2, p. 3-36.

Parenti C., 1999, Lockdown America: Police and Prisons in the Age of Crisis, New York, ed. Verso.

Perry W. L., McInnis B., Price C. C., Smith S. C. andt Holywood, J. S., 2013, « Predictive Policing: The Role of Crime Forecasting in Law Enforcement Operations », Rand Corporation Safety and Justice Program, p 1-189.

Ribeiro L., 2014, « Policiamento Comunitário », in Lima R, Ratton J. etAzevedo R. Crime (dir.), Polícia e Justiça no Brasil, São Paulo, éd. Contexto, p. 435-444.

Rouvroy A. and Berns T., 2018, « Governamentalidade Algorítmica e Perspectivas de emancipação: o díspar como condição de individuação pela relação? », in Bruno F., Cardoso B., Kanashiro M., Guilhon L. e Melgaço L. (dir.), Tecnopolíticas da Vigilância: Perspectivas da Margem, São Paulo, éd. Boitempo, p. 107-140.

Saap, 2016, « Projeto da SSAP orienta moradores a adotar medidas para manter o Alto dos Pinheiros seguro », Saap blog, [].

Sassen S., 2006, Territory, Authority, Rights: From Medieval to Global Assemblages, New Jersey, Princeton University Press.

Sinhoreto J., Schlittler M. C e Silvestre G, 2016, « Juventude e Violência Policial no Município de São Paulo », Revista Brasileira de Segurança Pública, n° 1, p. 10-35.

Soares J., 2017, « Moradores investem em circuitos eletrônicos para vigiar as ruas », in Veja S.P.,[].

Skolnic J. H. e Bayley D. H., 2006, Policiamento Comunitário, São Paulo, éd. Edusp.

Smart Cities Connect (SCC), 2017, « São Paulo Digitalizes Public Services », in Media & Research []

Tribunal de Contas do Estado, 2016, « Relatório de Fiscalização de Natureza Operacional », Diretoria de Contas do Governador.

Villela G., 2014, « Lei de 1941 considera ociosidade crime e pune “Vadiagem” com prisão de 3 meses », O Globo [].

Wacquant L., 2003, Punir os Pobres: A Nova Gestão da Miséria nos Estados Unidos, Rio de Janeiro, ed. Revan.

Williams M., 2016, « Global Security Assemblages », in Abrahamsen R. and Leander A. (eds), Routledge Handbook of Private Security Studies, New York, ed. Routledge, p. 131-139.

Haut de page


1 We realize that this diagnoses still open, and there are several critics on it. However, it fits as a strong description of the movements observable during the 90’s.

2 At this point is fundamental to mention the work of Clarke (1980), who also explores how this drive to a new environmental criminology prevailed. Contraposing the positivist criminology, the situational crime prevention focuses on the management and manipulation of the environment, in order to reduce the opportunity of crimes, while increases the possibility of its detection. Together with Clarke, several other authors also developed studies analyzing this new criminology, and gave situational prevention more theoretical density, such as Brantingham and Brantingham (1993), and Felson (2002).

3 About this, one of our interviewed, a coronel of the São Paulo’s Military Police responsible for the implementation of Detecta, attests the efficiency of the system by affirming that “through this tool, in a matter of seconds I can get all this kind of information (…) It’s really a Situational Awareness tool. This greatly helps the process of service and dispatch” (Secretaria S1, 2018).

4 According to Holston (2013), the “insurgent citizenship” is a perception that questions the regimes of privileges, expressing wishes and demands rights to access the city. The author points that even though the democracy has substantially evolved in Brazil during the past twenty years, it was followed by new types of violence, injustice, corruption, and impunity. In his words, “(…) at the same time that a generation of insurgent citizens democratized the urban space, creating an unpreceded access to its resources, an environment of fear and incivility permeated the public encounters. This new strangeness produces the abandon of the public space, the strengthen of the residences, the criminalization of the poor and the support of the police violence” (Holston, 2013, p. 349).

5 About this, the security and monitoring company Aster developed a study called “Janitoring together” about the neighborhood “Vila Leopoldina”, in São Paulo. In this study, they qualify the region as a highly risk place, once there is a strong socioeconomical contrast that leads to an intense circulation of people, irregular occupation, homeless people, prostitution side by side with middle class commercial spaces and residences. This diagnosis leads to a sort of solutions presented by the company that requires a janitorial attitude of the residents over public spaces, and the development of a private monitoring center, with an intense data interchange between private and public systems, as an effort to modulate the circulation of people in that neighborhood. They even recommend a public-private partnership between their systems and Detecta, sustaining that urban security is a resultant of an intensive surveillance program. The debased solution was, then, a “different plan of personal and patrimonial security”, that would include “advanced crime surveillance, inhibition and prevention processes and equipment”, a plan that could involve “both public security and private security”, and finally “technology of social articulation, reestablishing the quality of the living and minimizing occurrences” (Aster, 2017).

6 The term janitorial is an immediate translation from the term “zeladoria” in Portuguese, which encompasses ‘order’ and ‘security’ in a peculiarly way, which also clearly refers to the logic of the Gated Communities and, potentially, the role of the síndico (the janitor) – especially in how it articulates the public and the private. This is something we might want to look into more deeply, not least given the fact that the SegD’Boa home page reads “We offer you the quality and janitorial services that you are looking for”, while one of Aster’s projects – which also mentions Detecta and the ‘Suspicious Cam’ – is called “Together Janitoring Vila Leopoldina”. Also, it is significant to mention that ‘zelador’ derives from ‘zelar’ – “care for; surveill, protect, take care of something or someone with all care, attention and interest”.

7 As an example, some of the local companies that provides surveillance and algorithmic based cameras are related to transnational companies, or global alliances for security, such as the “Global Securalliance”, compromised in producing standardized practices in the exercise of private security, and consulting for public security agencies.

8 Regarding these “fears” and “demands”, an article recently published by Forum Magazine (2019) revealed that residents from upper middle class neighborhood “Higienópolis” demanded its “sanitization” by expelling homeless people, who were frequently described as thieves, vandals, or simply “human trash” in a semantic continuum similar to what Feldman (2004) understands as a common attempt to securitize groups who have no relationship with criminality or illegalisms, but somehow figure as a menace to the lifestyle of dominant ones, expanding the available mechanisms of control and sanction.

9 The real name was changed, once it is related to a third part.

Haut de page

Table des illustrations

Titre Graphic 1: Crimes Committed in upper middle-class and rich neighborhoods in São Paulo, as % of residents
Crédits Source: (Secretariat of Public Security of São Paulo, 2018)
Fichier image/jpeg, 26k
Titre Figure 1: Magazines Distributed in Electronic Security Fairs
Crédits Source: (Author’s collection, 2018, São Paulo)
Fichier image/jpeg, 274k
Titre Figure 2: Examples of public-private Security perimeters
Fichier image/jpeg, 178k
Titre Security perimeters sign
Crédits Source: (Author´s collection, 2018, São Paulo)
Fichier image/jpeg, 80k
Titre Figure 3: “Loitering” Analytic in São Paulo
Crédits Source: Extracted from the TV show “Mais Você”, 2017, São Paulo
Fichier image/png, 2,1M
Titre Figure 4: Videowall of the University`s Control Room
Crédits Source: (Author’s collection, 2018, São Paulo)
Fichier image/png, 323k
Haut de page

Pour citer cet article

Référence électronique

Alcides Eduardo dos Reis Peron et Marcos César Alvarez, « Governing the City: The Detecta Surveillance System in São Paulo and the Role of Private Vigilantism in the Public Security »Sciences et actions sociales [En ligne], 12 | 2019, mis en ligne le 17 décembre 2019, consulté le 21 juillet 2024. URL :

Haut de page


Alcides Eduardo dos Reis Peron

Postdoctoral Researcher – Department of Sociology – FFLCH/University of São Paulo,

Marcos César Alvarez

Professor at the Department of Sociology – FFLCH/University of São Paulo, Brazil
Vice Director of the Núcleo de Estudos da Violência – University of São Paulo

Haut de page

Droits d’auteur


Le texte seul est utilisable sous licence CC BY-NC-ND 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search