Apartheid drone: Infrastructures of militarism and the hidden genealogies of the South African Seeker

Drone infrastructure: Militarism embedded in everyday life

Violence executed by drone aircraft – also called unmanned systems or remotely piloted vehicles – is often theorized through the framework of targeting. On screen, an operator opens a kill box to attack a human target from the air platform. Chamayou (2015: 55) describes the kill box as a ‘a temporary autonomous zone of slaughter’ that could potentially be launched against an individual anywhere, at any time. This distant, yet immersive, view establishes a vertical field of sovereignty, defined as the supreme authority over a body politic, that expands and unbinds state power from territory (see Schmitt, 2006 [1950]). The power and perversity of this scope has framed and informed much of the recent theorizing about drone aircraft (Bousquet, 2018; Gusterson, 2016; Shaw, 2016; Wall and Monahan, 2011), leading scholars to articulate drone power as a machine optic that dehumanizes politics on a global scale, determining who lives and dies. US drone strikes over the past 20 year have systematically killed thousands, primarily in Somalia, Yemen, Pakistan, and Afghanistan. Rather than marking a new stage of technologically enabled war, however, drone aircraft actualize histories of targeting, technology, and airpower ordered not just by the war on terror but also by pre-existing racial, colonial, and Cold War divisions (Atanasoski and Vora, 2019; Chandler, 2020; Feldman, 2011; Galison, 1994; Kindervater, 2016; Masco, 2006; Mbembe, 2003; Satia, 2014). Moreover, a number of scholars studying Pakistan and Afghanistan have also pushed back against the optics of power articulated by drone theorists, drawing on the ‘grounded’ voices and resistances from the regions that are attacked (Durrani, 2016; Nath, 2017; Tahir, 2017). As I outline in this case study, the organization of state power through drone technology builds on patterns of domination that predate contemporary uses in the war on terror, fitting the drone’s lethal optic within a network of global exchange that intertwines colonialism, white supremacy, state power and military industry, increasingly rendered as ‘ordinary’ infrastructure.

Drone aircraft are not only defined by how networked systems are used to surveil and target, but also by what the drone hides and defends. Gregory’s (2017) update to his earlier essay emphasizes how, instead of being derived from a single optic, US drone strikes rely on a ‘deadly dance’ between competing and collaborating powers in the borderlands between Afghanistan and Pakistan, troubling their association with a single sovereign. Kaplan and Miller (2019) develop the concept of ‘atmospheric policing’ to describe how the use of drone aircraft creates a continuum of police and military power, drawing on a case study of the US-Mexico border. They write, ‘[drone aircraft] at every scale and range change and interact with atmospheres through which they move and mediate relations on the ground’ (p. 437). The relations created through drone infrastructure – both in the day-to-day use of the unmanned aircraft and in how the history of the drone is told – re-enact racial and colonial hierarchies through transnational networks and grounded relations that connect militarism to everyday life (Chow, 2006; Kaplan, 2018; Stewart, 2007). As Parks (2016: 232) writes, ‘Drone operations shape where people move and how they communicate, which buildings stand and which are destroyed, who shall live and who shall die. The drone is as much a technology of inscription as it is a technology of sensing or representation’.

This article examines the normalization of drone infrastructure in South Africa as it is transformed from an apartheid-era, secret weapon to a technology of national protection and environmental management. Its analytic builds on research that details how race is embedded in socio-technical systems, as a contingent set of relations that perpetuate racial categories and exclusions (see Atanasoski and Vora, 2019; Bowker and Star, 1999; Chun, 2008; Nakamura, 2002, 2007; Noble, 2018). To be clear, I am not arguing for a fixed relation between the drone and racial hierarchies, but examining how the Seeker organizes relations that privilege a particular population. In Society Must Be Defended, Foucault (2003: 15–16) observes that ‘power relations, as they function in a society like ours, are essentially anchored in a certain relationship of force that was established in and through war … the role of political power is perpetually to use a sort of silent war to reinscribe that relationship of force, and to reinscribe it in institutions, economic inequalities, and even in the bodies of individuals’. Yet, it is not only war that articulates these divisions, as Mbembe (2003) explains in his writing on necropower, but also the plantation and the colony. These relations have produced ‘new and unique forms of social existence in which vast populations are subjected to conditions of life conferring upon them the status of living dead’ (p. 39). Besteman (2019) uses the term ‘globalized military apartheid’ to describe how racialized exclusions, militarized borders and differential access to mobility shape the contemporary globe. Their counterparts, I argue, are unique zones of protection, exemplified by the privilege accorded to whites in apartheid-era South Africa, that use the logic of war and technology to naturalize how some populations live while others are left to die.

While my analysis is focused on the South African drone, I tie this research to recent writing on race, technology and surveillance, situating the Seeker’s history in a transnational context. Benjamin (2019) argues that race functions as a technology, ‘one designed to separate, stratify, and sanctify the many forms of injustice experienced by members of racialized groups’ (p. 36). Her argument draws on Hall (1997), who explains that ‘race as a principle of classification operates to sort out the world into its superiors and inferiors’ (p. 3). Hall’s analysis addresses the supposedly biological or genetic foundation of race. Benjamin expands, noting that technology reproduces and exacerbates ‘biological’ classifications to ‘hide, speed up, or reinforce racism’ (p. 36). Technology thus serves as a natural guise for divisions, rendering racial categories as implicit parts of socio-technical systems. Similarly, Jefferson’s (2020) work on algorithmic policing observes ‘an ongoing form of racialized social management that is slowly mutating through digital technology’ (p. 6), which relies on digital techniques to create zones of exclusion that target black and brown populations. Browne (2015) underlines connections between ‘how things get ordered racially’ and the history of surveillance, noting that, while these practices ‘[depend] on space and time and [are] subject to change’, they ‘most often [uphold] negating strategies that first accompanied European colonial expansion and transatlantic slavery that sought to structure social relations and institutions in ways that privilege whiteness’ (p. 17). Breckenridge (2014) ties the development of biometrics to apartheid, underlining how this history fits with new forms of global surveillance. Miller (2019) examines the invisibility of whiteness in their case study of drone aircraft in the American South: Without a history, the drone is rendered insensible, teaching viewers how ‘not to perceive’ connections between the US war on terror and policing. It is the drone’s function as an absence, Miller argues, that ties the system to white supremacy. Together, this scholarship exposes links between contemporary socio-technical systems and ongoing patterns of racialization. My analysis of the Seeker helps expand on these links, drawing on theorizations of infrastructure to think through the everyday embeddedness over time of drone technology and its modes of producing and reproducing racial exclusions.

Studies of infrastructure have long emphasized the relation between background and foreground to examine how power and affect are articulated through human and nonhuman networks (see Amin, 2014; Berlant, 2016; Guma, 2020; McCormack, 2017). In their groundbreaking work, Star and Ruhleder (1996) emphasize how ‘infrastructure appears only as a relational property’ (p. 113), countering an image of infrastructure as a system ‘that is built and maintained, and which then sinks into an invisible background’ (p. 112). The apartheid-era drone underlines how racial categories are embedded in the Seeker. Yet, the connection between race and unmanned aircraft is multiple and changing. During apartheid, drone aircraft reproduced the divides that they claimed to monitor – defending a protected white population and exposing ‘others’ as targets. More recent efforts to normalize unmanned aircraft for wildlife management in South Africa obscure the drone’s necropolitical basis (Krupar, 2013; Mbembe, 2003; Miller, 2019). Larkin (2013: 330) writes that infrastructure ‘is a categorizing moment. Taken thoughtfully, it comprises a cultural analytic that highlights the epistemological and political commitments involved in selecting what one sees as infrastructural (and thus causal) and what one leaves out’. Visibility and invisibility in apartheid-era drones operate in tandem with racial and colonial hierarchies. In the post-apartheid era, the Seeker reiterates enmity and division, albeit in the name of protecting democracy and, later, the environment.

While power is a key aspect of infrastructure studies, the term’s connection to the normalization of militarism is understudied (see, however, Cowen, 2014; Graham and Marvin, 2001; Martínez and Sirri, 2020). The etymology of infrastructure outlined in the Oxford English Dictionary cites an early use in 1951: ‘This new term “infrastructure” … denotes fixed military facilities such as airfields, base installations and transport systems’. As these examples make clear, infrastructure as it is commonly understood is intertwined with the history of war making in the 20th century – communications, medical advances, logistics, transportation, energy, and utilities are all part of strategies for military domination. Yet, these projects are also often put forward as means of development and progress, wherein ‘technological’ superiority is aligned with the protection of a certain population (Grewal, 2017; Terry, 2017). Foregrounding the drone’s optic – and its capacity for violence – obscures how the ‘kill box’ is just the most visible aspect of the racial and colonial divisions organized by the drone, which include an obscured, and privileged, field of protection as its counterpart. Division and war remain at the basis of drone infrastructure, even as political, social, and material exchanges that built the system aim to overlay white superiority with national defense and wildlife management.

Apartheid drones: Hiding international networks and revealing privileged populations

A cheeky entry from 1977 in the encyclopedia of air vehicles, Jane’s All the World’s Aircraft, optimistically promoted burgeoning uses of remotely piloted vehicles (RPVs) and their stealth history. Jane’s at that time noted that ‘the RPVs of today are so different in many cases that the 32 pages in which they are described have a ‘cloak and dagger’ quality to rival James Bond at his best’ (Taylor, 1977: 63). Included among the entries in the 1979 edition of Jane’s is the ND-100 Observer, a South African remotely piloted vehicle built by National Dynamics: ‘The Observer, low-cost multi-purpose mini-RPV has been built as a private venture for various military and civil applications, including real-time forest surveillance; aerial photography; search missions; flood, wildfire, and shoreline pollution detection and mapping; powerline, pipeline, and highway traffic observation and patrol; air sampling and storm research; fishing law enforcement and fishery spotting; and aerial survey’ (Taylor and Munson, 1979: 624). The interconnection between these apparently banal uses of drone aircraft corresponded with the entrenchment of apartheid in South Africa in the late 1970s and early 1980s and the use of remotely piloted vehicles by the South African Defence Force. These descriptions make visible a ‘civil’ side of the drone while omitting their ‘cloak and dagger’ use in neocolonial wars in Southern Africa.

In targeted killings, operators monitor the battlefield and attack targets on screen through real-time image transmission from an unmanned platform. Many scholars of this military system trace its origins to Israel in the late 1970s and early 1980s. While experiments with television targeting were carried out as early as World War II and were widely tested in the 1970s by various countries, Israel is typically described as the first military force to field what became today’s platform for targeted killings. During the Battle of Bekaa Valley in 1982, unmanned aircraft transmitted video footage to Israeli commanders, who then used the system to track mobile surface-to-air missiles operated by the opposing Lebanese and Syrian forces (Ehrhard, 2010; Grant, 2002; Wagner and Sloan, 1992). However, Zaloga (2011: 22) notes that, a year prior to Bekaa Valley, two Israeli firms ‘offered a method to monitor the battlefield using flexible, low-cost air vehicles that could be operated easily from near the frontlines. This concept was first demonstrated in 1981 when the South African Army used the IAI (Israel Aircraft Industries) Scout during Operation Protea in Angola’. These drones were not yet armed; rather, they provided real-time images to a television monitor in a nearby ground station that were used for reconnaissance and to organize aerial attacks.

The South African Army’s 1981 demonstration is often treated as a footnote in the history of drone aircraft, which typically follows the connection between the Israeli drones used in 1982 and the Predator platform developed for US targeted killings in the war on terror (see Grant, 2002; Whittle, 2014). But the genealogy of the drone could instead follow the Israeli drone to South Africa, where the South African Defence Force bought and developed unmanned platforms to advance its military goals throughout the 1980s. The remotely piloted aircraft were used in the South African Border Wars and flown against African revolutionary forces operating in Angola and Mozambique. A Central Intelligence Agency (1986) report on unmanned aircraft observed the significance of the South African drone industry and its advances in weaponizing the drone: ‘National Dynamics, Ltd., of South Africa is developing the Eyrie multi-mission RPV capable of firing 2.75 mm rockets’ (p. 5). These aircraft, not just the drones flown in Bekaa Valley, presage targeted killings by the US. The report also is a reminder of secret US support for the South African Defence Force in their wars against anti-colonial revolutionaries who were allied with the Soviet Union during the 1980s (see Davies, 2007; Rich, 1988).

Decades later, the political and economic exchanges that made the South African Seeker have been effaced, leaving behind an altered narrative in which the Seeker is a uniquely South African creation. These moves are at work in an online forum, begun in 2014, devoted to South African ‘Secret projects’. I cite these discussions to examine how they position the aircraft with respect to apartheid, not to corroborate historical details. The opening post in the forum observes: ‘Unknown to many is that South Africa was one of the early (modern day) pioneers of RPV’s/UAV’s/Drones in terms of military use, later in 1994 we were the first to use them in commercial airspace for civilian purposes (monitoring the first open general elections). Early history on development is difficult to find, however this thread will hopefully uncover some of it and the associated developments’ (Secret Projects, n.d.). The author positions South Africa as a ‘pioneer’ of RPV/UAV technology and asserts a ‘we’ that aligns the nation, the drone and technological innovation. As is clear from the post, this ‘we’ is not just a generic reference to the nation at large but rather to the white population that created and benefited from the apartheid regime and enforced it with violence. Reference to the Seeker invokes exclusions that underwrite racial hierarchy.

Simply claiming that the history of these projects ‘is difficult to find’, the author elides the global exchange of military technologies that made the Seeker system. As recounted above, the drone was initially bought from Israel. The post notes, however, that the first RPV was built by the South African Council for Scientific and Industrial Research (CSIR) and four prototypes were tested by the South African military in 1978, referring to the National Dynamics project. The author then explains:

It seems then that some interaction with Chile followed, then shortly after that with Israel, from which we bought 5 x IAI Scout UAV’s - named ‘Cobalt’ in SAAF service (some say 13). Shortly thereafter we developed our own Kentron Seeker 2B and later 2C UAV’s. We started using them operationally during our own Bush-War in Angola and SWA/Namibia. (Secret Projects, n.d.)

I situate this comment in the broader context of South African military industry during apartheid. As mentioned at the outset, the circumstances surrounding the sale of the Scout from Israeli Aircraft Industries to South Africa are unclear.

In Apartheid, Guns and Money, van Vuuren (2017) investigates the web of illegal exchanges that provided armaments for the South African Defence Force after the 1977 arms embargo, linking material support for the apartheid regime to scores of nations across the globe. Illicit financial dealings and arms sales implicate more than forty countries globally, including the United States, France, the United Kingdom, the Soviet Union and China – all members of the United Nations Security Council. Polakow-Suransky explains in The Unspoken Alliance: Israel’s Secret Relationship with Apartheid South Africa (2010), however, that the connection between Israel and South Africa was particularly significant. He writes, ‘Israel profited handsomely from arms exports and South Africa gained access to cutting edge weaponry at a time when the rest of the world was turning against the apartheid state’ (p. 6). South African and Israeli weapons industries, both directly tied to the state, agreed to share technological developments with each other to bolster production in both nations. According to Armscor documents quoted in van Vuuren’s (2017: 471) book, ‘it was agreed [between Israel and South Africa] that no party would make manufacturing licenses or rights and skills in respect of defense or armaments available to other organizations or countries without the approval of the other party’. Weapons sales rely on end-user certificates to verify that the recipient was not planning on transferring the military technologies to another party. Van Vuuren further describes how ‘Israeli companies working in the armaments sector, such as Israeli Aircraft Industries (IAI), exported weapons to a third country that ‘may or may not be Indonesia’, and then Indonesia would allow the re-export of weapons to South Africa’ (p. 475). Other countries that facilitated the re-exportation of weapons to South Africa included Chile and Argentina – both dictatorships at the time, and with whom Israel also had substantial dealings. These hidden exchanges are all part of the backdrop to the South African Seeker.

Due to the secret nature of Israeli and South African cooperation, it is difficult to determine exactly how Israeli Scouts were first used in Operation Protea. ⁸ An article in African Defence Review explains, ‘[F]acing a growing war in Angola, the South African Air Force acquired the first five of what would eventually become a fleet of around over a dozen Israeli Aircraft Industries Scout UAVs of various marks and differing engine options in 1980. The aircraft were all given RPV designations in SAAF service, ranging from RPV-1B for the first five aircraft to RPV-2B for the IAI Scout 800s delivered in 1984 and project names included Gharra, Cobalt and Leghorn’ (Olivier, 2015). While the details are not available, it is possible to review how drone aircraft may have assisted South African forces in the air battle, however. Operation Protea aimed to destroy the South West Africa People’s Organization (SWAPO) command and training bases in present-day Angola. On August 21, 1981, just prior to the ground invasion, the South African Air Force destroyed air defense installations at and around the towns of Cahama and Chibemba in Angola, in what was known as Operation Konyn. This mission bears a conspicuous resemblance to the Israeli strike in Bekaa Valley, which, as noted above, is typically taken as the inauguration of today’s drone warfare. In the Bekaa Valley strike, drones were used to monitor and target air defense gunnery on the ground. The aircraft acted as a decoy to be attacked: By drawing fire from air defense systems on the ground, the unmanned platform revealed the location of surface-to-air defenses, which were subsequently attacked by aerial bombing (Grant, 2002).

Retrospective accounts of the once-invisible air platform build an implicit association between racial hierarchy, technological superiority and the drone decoy. Posts on ‘Secret projects’ highlight how the South African Defence Force used the Seeker aircraft in Mozambique and Angola against on-the-ground air defenses. Their description as ‘unmanned’ technologies suggests a kind of invincibility and invisible advantage over the enemy, who may have mistaken the RPV for manned aircraft. Tellingly, the online accounts of drone technologies emphasize the superiority of these weapons, even when they fail. One author in the forum recalls: ‘In 1983 a UAV was lost over Mozambique after it was shot down by a Soviet SA-3 missile, after quite a few SA-3 missile had already been fired at it, it was busy doing post strike damage assessment on a raid by 4 X Impala MK2 light strike aircraft, as a reprisal for the Church Street bombing in Pretoria’. Another post notes, ‘The book “Jam Stealer” by Peet Coetzee describes in 1983 how he witnessed the controllers duck as they watched a Sam (9?) narrowly miss an RPV that was launched at it while they were attempting to find the location of the launcher near Chibemba in southern Angola in order to capture it intact’ (The SAAF Forum, n.d.).

Another thread in the online forum focuses on a later use, also in Angola, during the last phase of military operations in 1987. It describes how drone aircraft drew anti-aircraft fire, probably by mimicking the radar signatures of piloted aircraft. A commentator in the forum writes, ‘It was amusing to read that when one [drone] was shot down, the enemy spent 3 days looking for the pilot. It was also amazing how far ahead in technology we were in those days, unfortunately I was in the infantry and we were always kept in the dark’ (The SAAF Forum, n.d.). The authors in the forum use the Seeker to advance claims of technological superiority, while the war against the African National Congress and neighboring African states is made invisible. The ‘secret’ drone invokes the protection of white South Africans and the exclusion of Black Africans. The posts attend to how the ‘unmanned’ aircraft are shot down during battles between ground and air forces, even though the operators are protected from a distance and, as a result, not actually under attack. The drone decoy is an invisible field of protection aligned with apartheid. The final post names a ‘we’ who is amused by an enemy who confuses piloted and pilotless aircraft, implying ‘their’ backwardness against ‘our’ technological advance.

The slippage between protection and superiority is further reflected in the transition of the drone from a reconnaissance and decoy system to an attack weapon. In the 1980s, South Africa expanded its UAV program to test its use as a missile platform. A 2001 report explains: ‘Weaponization began in cooperation with ARMSCOR, the state-owned arms producer, which developed unmanned aerial vehicles (UAVs) and missiles that would have the capability to carry chemical and biological agents’ (Burgess and Purkitt, 2001: 26). This echoes the observations of the CIA report quoted previously. On January 16, 1992, a chemical weapons attack was reportedly launched in Mozambique: ‘A gas similar to teargas, causing pain and irritation was allegedly sprayed from reconnaissance aircraft’ (p. 48). A later UN investigation into the incident was inconclusive: ‘These outside investigators had a great deal of difficulty disentangling the proximate causes of death since many of the victims were suffering from malnutrition and other diseases at the time of their death’ (p. 48). During the Truth and Reconciliation Commission, however, various testimonies alluded to chemical attacks by drone. One scientist recalled that a ‘drone aircraft with yellow smoke was tested after De Klerk’s speech against secret projects in 1990’ (in Burgess and Purkitt, 2001: 49). An ANC defense expert claimed, ‘The use of CBW [Chemical and Biological Weapons] against Mozambican troops preceded the whites only referendum in 1992 and was used to demonstrate the resolve of the de Klerk government’ (in Burgess and Purkitt, 2001: 49).

If these accounts are accurate, they would change the history of killing by drone aircraft, which has typically been associated with the US-led war on terror and its use of Predator aircraft in Afghanistan beginning in 2002. An Israeli-led mission to assassinate Hezbollah leader Abbas al-Moussawi on February 16, 1992, which used a drone and a helicopter to achieve the same effect as today’s unmanned aircraft platforms, is another recent orgin (see Chandler, 2020). However, it may be that the first attack mission by drone was conducted by South African forces in Mozambique, tying the history of targeted killing to apartheid and white supremacy and relocating this geneology to South Africa. Even if these incidents of South African use of unmanned weapons are apocryphal, they suggest a realignment of how we think about drone warfare and the infrastructure that makes it both visible and invisible. Aerial protection enacts racial exclusion and colonial hierarchy. Indeed, Mbembe (2003: 29) points out how UAVs are among an array of technologies used for occupation from the air: ‘Such precision is combined with the tactics of medieval siege warfare adapted to the networked sprawl of urban refugee camps. An orchestrated and systematic sabotage of the enemy’s societal and urban infrastructure network complements the appropriation of land, water, and airspace resources’. Mbembe’s reference to infrastructure to explain the ‘precision’ of UAVs is not incidental here. Infrastructure underlines the interconnectedness of the ground and sky in necropolitcs: The drone erases the enemy target, while it privileges a protected population. The system is organized to let some live and leave others to die. In South Africa, the vertical power claimed by drones is tied to the territory and settlers below. Yet division is not the only significant effect of drone infrastructure. The term also suggest how power organized through the military weapon can be refashioned for ostensibly non-military uses, shaping ‘civil’ life, and becoming part of the environment.

Democratic drones and wildlife protection: Revealing national protections and hiding divisions

Published following the independence of Namibia and the withdrawal of South African forces from Southern Africa, Campbell’s 1990 paper, ‘The dismantling of the apartheid war machine and the problems of the conversion of the military industrial complex’, analyzed the interconnections between the military, state racism and capital accumulation in South Africa. Campbell underlines the substantial challenges of refashioning the South African military post-apartheid: ‘Those who bore the brunt of the destruction imposed by militarization of the region were seeking peace and democracy. However, the articulation of how this peace could be secured was still elusive given the vast economic resources which had been deployed into the military’ (p. 1). Armscor takes a significant place in this analysis; Campbell notes, ‘Despite the secrecy which surrounds the procurement and production of arms in South Africa the reality is that ARMSCOR consumes resources that could be used to accelerate social development and economic reconstruction’ (p. 15). He reports how national security during apartheid extended to all aspects of government and brought together political, military, economic, scientific and religious-cultural actions. Campbell’s conclusion explains, ‘Liberation is not simply a military/technical issue but a question linked to the consciousness and political mobilization of social forces with an interest in democratic mobilization and participation’ (p. 21). The trajectory of South Africa’s unmanned aircraft, the Seeker, conveys how military-industry refashioned itself post-apartheid, as well as the limits of these transformations.

Beginning in the 1990s, Armscor aimed to reposition the wartime Seeker as a platform that could be used in the service of the newly democratic nation. This change is consistent with transformations to the South African military, which in 1994 integrated the former South African Defence Forces (SADF) with revolutionary freedom fighters from uMkhonto we Sizwe, the liberation army of the African National Congress, and the Azanian People’s Liberation Army of the Pan African Congress, in addition to four homeland armies. The incorporation of these military forces ‘was one of the most far reaching and complex undertakings in public reform in post-apartheid South Africa’, bringing former wartime enemies together to create the South African National Defence Forces (see Heinecken and van der Waag-Cowling, 2009: 518). The Seeker technology was similarly rebranded as a tool for democracy, though the touted new uses of the drone did not become significant in South Africa. The shifts from 1994 onward highlight how threat and protection are refashioned by drone infrastructure to both extend and obscure the racial exclusions of apartheid. The drone is remade as a democratic tool for the new South African nation. The racial and colonial hierarchy that divided the Seeker and an enemy target is reorganized as border, police, and environmental control in the post-apartheid era. The once-secret air platform is made visible as a technological proxy for democratic progress or international development, while targeting – which remains inherent in these new uses – takes the guise of domestic and environmental protections.

Importantly, these changes were not linear. Globally, drone aircraft continue to explicitly reiterate racial and colonial hierarchies as a secret weapon: The US Predator drone took up the classified project of targeted killing in the Middle East, while Israeli forces executed a program of targeted killing against Palestinians during the Second Intifada (see Hajjar, 2017). And though South African drones have been repurposed for ‘peacetime’ uses, they continue to be built and exported for military uses as well. Between 1996 and 2010, Armscor (later, Denel) sold the remotely piloted system to Algeria, the United Arab Emirates, and Uganda. The changes to the Seeker in the 1990s are instructive, as visibility and concealment organize affective and material relations. No longer secret, the Seeker is a machine proxy for the democratic nation, a tool of wildlife management and international development. In both military and nonmilitary applications in South Africa, the system is positioned as an apolitical infrastructure and its history is erased.

These changes are evident in public records associated with the Seeker. With the transition to democracy, previously guarded secrets of weapons procurement and manufacturing became matters for corporate recordkeeping. Armscor, as outlined above, was central to the South Africa arms industry during apartheid – both building weapons and purchasing them illicitly. The first available annual report for the company, published in 1992–1993, explains: ‘Such a report has not been possible given the security situation surrounding the corporation’s activities’ (p. 2). That apartheid was the basis for the corporation’s ‘security situation’ is omitted. Rather, the document positions Armscor as ‘an apolitical organization which serves the government of the day’ (p. i). Racial and colonial hierarchy is replaced with the ostensibly neutral goal of government service. Significantly, the only history that is mentioned in the annual report is the company’s role in conservation: ‘Armscor has a particularly proud record in the field of environmental management. Throughout Armscor’s existence the corporation’s management has been extremely aware of its responsibility to the environment and has done everything possible to ensure that its activities would have no detrimental impact on the world around it’ (p. 20). Line drawings of leopards appear throughout the report, linking weapons procurement and development with images of endangered species native to South Africa.

The corporation’s overlay of democracy, weapons procurement and environment continues in the next year as well, calling attention to its policies that focus on its commitments to the people of South Africa. Its 1993–1994 report begins, ‘The national and international environment is changing rapidly. So is Armscor’ (p. 3). The opening pages promote the company’s commitment to ‘greater transparency’ to ‘serve the people of South Africa at large and especially the local communities with whom it is involved’ (p. 3). Implicit in this sentence is a recognition that weapons production during apartheid served a privileged, white community; the reformed company aims broaden its service to the ‘people of South Africa’. The first time Armscor explicitly mentions the Seeker unmanned aircraft is in the 1994–1995 report, which reveals the Seeker’s role in monitoring the election of Nelson Mandela and patrolling South Africa’s borders:

The Seeker remotely piloted vehicle (an unmanned aircraft), developed some years ago, was successfully deployed during the 1994 elections to monitor unrest areas. The SA Army is also using the system to monitor attempts to cross the national borders illegally. Their use of the system paid off in the Eastern Transvaal where movements of illegal aliens and gun-runners were detected and suspects apprehended. (p. 10)

In this passage, the Seeker is tied to the 1994 elections and positioned against ‘unrest’. The unmanned aircraft is made visible as a tool of ‘successful’ governance and aligned with a peaceful transition to democratic power. The association between the Seeker and national protection continues in the next sentence, depicting drone infrastructure as countering ‘illegal aliens’ and ‘gun-runners’. Here, the Seeker is aligned with policing the national border, as a tool used to apprehend ‘suspects’.

The applications of the Seeker for ‘peacetime’ policing are further elaborated in Fourie’s report, ‘Operations with Seeker in Southern Africa’ (Shephard Conferences & Exhibitions, 1999), discussed at the outset of this article, which was presented in conjunction with the Paris Air Show. The report aimed to represent a reformed South African National Defence Force, even as it served to promote global sales of the South African unmanned platform. It describes the unmanned aircraft as a ‘Flagship of the South African National Defence Force’ and explains how the system was transformed in the 1990s to support the South African Police Services and other state security departments. The 1999 report lists the following ‘successes’ of the Seeker in newly democratic South Africa:

Apprehension of over 400 illegal immigrants from Mozambique.

Apprehension of illegal immigrants from Zimbabwe.

Arrests of feared tribal warlords in KwaZulu/Natal, who have been evading the police for more than 9 year.

Arrests of gang leaders involved in gun running and possession of illegal/home made[sic] weapons.

Arrests of cattle and livestock thieves (cross-border livestock thieving on the Mozambique and Swaziland borders).

Location of aircraft wreckages in areas that search aircraft were unable to access.

These operations propose a new set of oppositions as the basis of drone infrastructure. Instead of an enemy target, the system surveils ‘illegal immigrants’, ‘feared tribal warlords’, ‘gang leaders’, and ‘cattle and livestock thieves’. The drone acts to apprehend and arrest persons that the state regards as criminal.

In the 1990s, mobilization of the South African Seeker by the military, government, and police reversed earlier patterns of concealment and exposure to emphasize its role in safeguarding the nation state. Rather than countering an enemy target, Armscor promoted the unmanned platform as a technological stand-in for police protections. Accounts aimed to neutralize the system as an apolitical technology, to erase its apartheid-era history. As a military weapon, the Seeker was primarily deployed during the border wars with Angola and Mozambique. Militant targets are replaced by the ‘apprehension of over 400 illegal immigrants from Mozambique’. The KwaZulu-Natal region, also mentioned above, had been a political focal point for the African National Congress (ANC) and anti-apartheid struggle, while the Kingdom of Swaziland was a political refuge and military training site for leaders of the ANC. These examples show the overlaps between the populations and regions – if not the individuals themselves – targeted by the apartheid-era Seeker and those promoted as stabilizing democratic South Africa. The Seeker continued to target and divide by shoring up the boundaries of the state and protecting it against purportedly ‘criminal’ individuals and groups. This shift could also be seen in the crews contracted to operate the Seeker: Fourie’s report indicates that, while the military squadron associated with the system closed in 1991, the Seeker continued to be flown by a technical crew contracted from the manufacturer and by members of the South African Air Force. Although the Seeker no longer corresponded with the explicit hierarchies of apartheid, peacetime missions nonetheless repeated the Seeker’s wartime flights in the name of democracy and technological advance.

The 1999 report also observes, ‘During political rallies, mass funeral marches and election proceedings, Seeker was utilized to assist the South African Police Services in maintaining peace and stability during these events. The mere presence of Seeker and the knowledge that the entire process was being monitored, ensured that no unrest and disputes occurred’. Earlier accounts of the Seeker described how the platform acted against an enemy target, providing a ‘secret’ advantage for South African military forces. When the system was transferred to the South African Police Service, the Seeker became a ‘presence’ that monitored populations and enforced stability. Both uses of the Seeker create an infrastructure that divides between above and below – the first in the name of overt superiority and the second to maintain stability. In the operations in the 1990s, the ‘peacetime’ missions replaced military advance with a technological safeguard. As with the missions for border protection and crime, the so-called new uses of the Seeker in democratic South Africa continued to reinforce apartheid divides. Political rallies, mass funeral marches and election proceedings represent a Black majority population. Left unsaid in the Armscor documents are the military and police force’s role in creating the very divides that the Seeker is supposed to stabilize.

‘Operations with Seeker in Southern Africa’ not only describes the drone as protecting the nation against illegal immigrants, criminals, gangs and mass unrest, but also suggests how the unmanned platform might be applied to environmental protection and infrastructure projects. Echoing the uses described in the 1979 Jane’s entry on the South African Observer, Fourie’s 1999 report explains, ‘Seeker is also utilized by semi- and non-governmental departments in assistance with various causes. The clients for these tasks include the Forestry Department, National Parks Board, Harbours and Fisheries, Tecom and Escom’ (p. 9). In particular, the paper notes the Seeker’s use in ‘[g]ame counting and assistance with location of game poachers for the National Parks Board’ (p. 9). In the 1997–1998 Armscor annual report, environmental protection takes up several pages. The final sentence explains, ‘ARMSCOR is looking for long-term survival in the environment where it carries out its task, and wishes to be recognized by the community as an organization that cares about the environment. ARMSCOR, which is also a corporate partner of the World Wildlife Fund, is a reliable partner to those who are environmentally aware’ (p. 47).

In the following decade, 2000–2010, military and policing uses for the Seeker in South Africa increasingly receded to the background as Armscor was dissolved and the manufacturing portfolio for unmanned aircraft was restructured to become part of Denel, a new state-owned military corporation created by South African ANC leadership. ⁹ Against the rising prominence of US military drones, the Seeker becomes a footnote to these aircraft, primarily associated with civilian applications, despite continued development of military-grade unmanned aircraft by South Africa (see Airforce Technology, 2014). For example, the final sentences of a 2003 article in The Wall Street Journal describes the South African Seeker as having transitioned to non-military purposes, noting, ‘Among its uses: locating poachers in game reserves’ (Stone, 2003). In 2010, the Council for Scientific and Industrial Research, which had initially developed the drone project decades earlier, was ‘tasked with leading the countries [unmanned aircraft systems] research and development for civilian applications’ (p. 29).

The most recent uses of the drone aircraft in South Africa suggest an infrastructure that protects endangered wildlife, associating the unmanned platform with international goals of conservation. Operations with unmanned aircraft in South Africa are described as a response to the rhinoceros poaching crisis in the 2000s, fueled by the illicit trade of their horn to Asia. Recent projects are presented as disconnected from the long history of unmanned aircraft development in South Africa. Rather, unmanned platforms are connected to international environmental organizations such as the World Wildlife Fund, transnational corporations such as DJI, the current market leader in easy-to-fly drones and aerial photography systems based in China, and African Parks, a conservation organization that manages nineteen national parks and protected areas in eleven countries in Africa. In public discourse, these uses are often framed as positive outcomes of technological innovation, falling under the heading of ‘drones for good’. A New York Times headline from 2017 exemplifies drone infrastructure in South Africa today: ‘High above, drones keep watchful eyes on wildlife in Africa’ (Nuwer, 2017). The article promotes unmanned systems as a ‘high-tech’ solution for wildlife management. Similarly, an article published in The Guardian in 2016 titled, ‘From killing machines to agents of hope: The future of drones in Africa’, proposes a seamless transition for unmanned aircraft from targeted killings by US drones in the Middle East to their use for humanitarian goals in Africa (Flood, 2016). Drone aircraft in South Africa are this time made visible as an international, technological solution for a conservation crisis.

The use of drones to survey and protect endangered animal species draws upon the entrenched racial and colonial histories that shaped the Seeker, however. The drones’ protection of the rhinoceros and targeting of poachers is underwritten by transnational exchanges of capital and aid. Drones deployed in national parks in South Africa aim to ‘save’ animal populations. They counter the longstanding stereotype of a Black African poacher. Monitored by drone, the endangered animal takes on the role of a protected population while poachers become the targeted enemy. These contemporary relations draw on the alignment of environmental protection, nation-building and drone innovation found in reports on weapons development and procurement in the earlier apartheid and post-apartheid era. ¹⁰ The prominence of the drone in these accounts underlines the militarization of wildlife conservation over the past decade, even if the use of unmanned platforms is described as a non-military application. South African rangers are led by former military personnel and deploy a range of high- and low-tech solutions to counter poachers, which include not only drone aircraft but also support helicopters, ground patrols and tracking dogs (Ramsay, 2014). It is unclear how many poachers have been arrested or killed through these patrols – South African officials have not released official numbers – but data from organized crime statistics indicate 150–200 poachers killed between 2010 and 2015. Joaquim Chissano, the former president of Mozambique, has accused South African rangers of killing 476 Mozambiquan poachers during this period (O’Grady, 2020).

While drone optics appear to iterate aerial power for conservation, the infrastructure of the drone has thus far not functioned to advance anti-poaching. In a video accompanying the article in The Guardian (Flood, 2016), a team of multi-racial drone operators demonstrates the technology, while the voice-over observes, ‘The operators say they have seen a direct link between their presence in the area and a fall in poaching’. The film cuts to an unmanned aircraft hovering over the savannah. Yet in 2017, when the projects were reviewed by officials from South African National Parks, they found that the year-long experiment to use unmanned aircraft to detect poachers revealed serious shortcomings. Due to their small size and weight, the unmanned platforms tested were not able to carry sophisticated payloads and, as a result, the experiment did not lead to an apprehension of poachers during the trial period (Martin, 2017). More importantly, solutions for poaching have to account for the increased demand in Asia for rhinoceros’ horn, as well as the social, political, and economic conditions of the communities around the protected areas. Poaching is not a ‘new’ problem that can simply be countered by technological advance; rather, it marks the persistence of unaddressed racial and economic inequalities in a global context. Like other infrastructures, drone aircraft can fail and, in this case, are inadequate to the problem they are supposed to solve.

Conclusion

Given its history, the term ‘infrastructure’ articulates how militarism is interwoven with everyday life. The case of the South African Seeker shows how drone aircraft were advanced by military-industry and global networks of exchange that supported racist protections. Yet the account is also a cautionary tale for the present moment. As the novelty of drone warfare recedes, the framework of protection and targeting is reiterated as a part of international aid and commerce. A 2017 report states, ‘Commercial drones are here: The future of unmanned aerial systems (UAS)’, released by the global management consulting firm McKinsey & Company (2017), indicates that drone start-up companies have attracted more than 3 billion dollars in funding to ‘explore new UAS applications, industries, and geographies’. Like Fourie’s 1999 ‘Operations with Seeker in Southern Africa’, the report points to the potential of non-military drone applications. These are described as an opportunity for commercial gains, worth up to $46 billion by 2026. This future, however, will be interwoven with the militarism of drone infrastructure. The Seeker challenges a straightforward shift from killing to saving promoted by drone advocates and instead indicates the persistence of violence in drone infrastructure and its role in perpetuating racial and colonial histories of power.

Charting how the Seeker is made and unmade, I also unsettle a simple characterization of the drone as a mere tool or product. Drone infrastructure, as I have developed the term, describes relations of power as they move and transform in time and space. I complement analyses that have emphasized the optics of vertical sovereignty with a case study of the Seeker that describes drone infrastructure as a genealogy of ‘categorizing moments’ (Larkin, 2013). The Seeker is created and recreated through the interplay between air power and territory; what the system targets and hides; and who it privileges and protects. The systems, moreover, fail in their totalizing aims, proving inadequate to the military, democratic and environmental goals that the drone is supposed to advance. Guma (2022) insists on the incompleteness of infrastructure to counter normative descriptions of wholeness and completeness. He writes, ‘Infrastructure development thus is better construed as a continuous preoccupation that includes patching up, reconfiguring, and reassembling processes that allow for their incremental and continual redefinition’ (p. 5). Given the trajectories of drone infrastructure outlined above and recognizing the enduring military logics of racialized exclusion that underlie its supposedly neutral or benevolent uses, its inadequacies might also become groundwork to contest militarism’s everyday embeddedness in South Africa and elsewhere.