Extraoperable networks: Delay and power in the rise of LunaNet

Abstract

This article takes us to the Moon where LunaNet, a communication network designed to support NASA’s Artemis program, is emerging to enable sustained human presence in space. My specific focus is on the LunaNet’s interoperability discourse which situates it within the technical, historical, and political trajectories of the Internet and within socio‑technical structures that shape power here on Earth. By analyzing this discourse, I argue that, when applied to outer space, the visions of interoperability are limited, overlooking what makes LunaNet distinctive—its ability to function without continuous connectivity. To foreground this point, I introduce the concept of extraoperability, a speculative framework that challenges Earth‑bound assumptions about Internet connectivity and proposes that alternative modes of relation may emerge when we network beyond our Blue Planet.

Introduction

An astronaut on the southern pole of the Moon, wearing the colors of the American flag, crouches to reach a cable plugged into a socket in the lunar regolith. Seen from the Moon’s surface, next to a satellite, our planet is a small gentle dot bleeding blue into the darkness of the sky. The second astronaut holds a tablet, an iPad lookalike. It provides real-time weather information, which is part of the astronauts’ lifeline since the Moon’s surface is too extreme for human life without technological aids. The screen shows the surface temperature at −210°C and the wind speed at 421 km/s. The tablet is connected to NASA’s LunaNet, and its network status is online.

The preceding depiction of the astronauts is an artist’s rendering published as part of NASA’s press release titled “LunaNet: Empowering Artemis with Communications and Navigation Interoperability for Space Communication” (Manning, 2023) and appearing on a NASA video “LunaNet: Enhancing Connectivity and Empowering Missions at the Moon” (NASA’s Goddard Space Flight Center, 2021). The future space missions will be more connected than ever, the LunaNet video explains, and describes how on the Moon, the network can be used for communication, navigation, and to alert astronauts of approaching solar storms so they can seek shelter before it is too late. “LunaNet directly supports NASA’s Artemis program,” the video tells us (NASA’s Goddard Space Flight Center, 2021). The goal of the Artemis program is to establish humanity’s “long-term presence at the Moon” (Schauer and Baird, 2021). As part of this bigger vision, the LunaNet is positioned to form a backbone communication network for the Moon, Mars and the perhaps the entire solar system.

Although space, as the environment in which LunaNet operates, differs greatly from what we are accustomed to here on Earth, the network itself is often described as resembling our terrestrial Internet. “LunaNet offers a network approach similar to the internet on Earth, where users maintain connections with the larger network and do not need to schedule data transference in advance,” NASA explains (Schauer and Baird, 2021). The network’s instrumental value appears self-explanatory: “Astronauts and space tourists are seen to have ready Internet access once they land on the moon” (Palmers, 2021).

We might want to make legible the role of the Internet in this discourse. The access to the Internet gives our world a particular shape. It is a gateway to the cultural, political, and economic flows of order here on Earth. This was the argument of critical Internet studies of the early 2000s: models of connectivity—and the specificities that characterize them—profoundly structure how we come together (cf. Chun, 2008; Galloway, 2004; Galloway and Thacker, 2013). In this vein, it is essential to note that while the LunaNet is compared to and interoperable with the Internet, it is also different in a fundamental technical sense (Israel et al., 2020). The way that the Internet connects nodes here on Earth does not work further we go to space: the transmission delays between nodes that are positioned in different parts of our solar system are too long, and the signals between them too unstable for continuous end-to-end connectivity (DeNardis, 2023; Israel et al., 2020: 4). For this reason, the Artemis mission needs something like LunaNet from the outset.

The task of this article is to critically examine the interoperability discourse of the LunaNet (NASA, 2025; Schauer and Baird, 2021) as part of plans to design information networks for outer space (DeNardis, 2023). I analyze talks from the Lunar Interoperability Forum hosted by the White House in May 2024, an episode of NASA’s (2021) The Invisible Network podcast from May 19, 2021 produced by NASA’s SCaN program, and other public commentaries by the people involved in different capacities in the LunaNet project. I also explore the network’s technical documentation, including the LunaNet Interoperability Specification Document (NASA, 2025; see also Manning. 2023) developed collaboratively by NASA, the European Space Agency (ESA), and the Japan Aerospace Exploration Agency (JAXA), and which defines the standards, protocols, and interfaces that apply to this network. My claim is that interoperability is not a neutral condition that simply enables networks, individuals, and other entities to connect; what is given instead is that specific forms of power necessitate interoperability.

The emphasis of this article is in developing media theory for networks beyond Earth rather than doing a methodological exercise in discourse analysis. For analytical purposes, the article is divided into two main sections: here and there. In the “Here” section, here refers to Earth. I present how experts and researchers discuss LunaNet and position the network in the context and continuum of Internet history and the modes of power that it entails. In the “There” section, I develop the notion of extraoperability by applying Friedrich Kittler’s (1999, 2006, 2022) media theory in conjunction with Luciana Parisi’s (2013) and Isabelle Stengers’ (2011a) Whitehead-inspired process philosophy. Extraoperability is used to explain how outer space, by necessitating a rethinking of how networking functions, also opens possibilities to challenge unified and dominant cultural, political, and economic ways of living.

Here

At the Lunar Interoperability Forum, Vinton Cerf—one of the designers of the Internet’s Transmission Control Protocol/Internet Protocol (TCP/IP) suite, Google’s Vice President and Chief Internet Evangelist, and one of the central minds behind extending the Internet to outer space—delivered a keynote address that mapped the development of interplanetary backbone networks for communications (Cerf, 2024). Interoperability is a central theme here, and Cerf (2024) wishes that everyone recognizes its importance:

I hope every single person in the room shares these aspirations with me. One of them is to create [. . .] an interoperable communication system that can be implemented by multiple parties in the same way that the public internet is implemented—many different groups participating. This should be similar in spirit to that. ¹

The forum is hosted by the Consultative Committee for Space Data Systems (CCSDS) and the Interagency Operations Advisory Group (IOAG), both international consortia that work on standardizing space missions and promoting interoperability among space agencies. The conference aims to accelerate “the development of all needed international space communication and navigation standards required to ensure multi-national and multi-system interoperability around the Moon” (Consultative Committee for Space Data Systems (CCSDS), n.d.).

In his talk, Cerf characterizes the development of a Solar System Internet, a general network architecture that NASA, in collaboration with others, has gradually built since the end of the 1990s. This model is grounded in a novel network architecture called Delay/Disruption Tolerant Networking (DTN) whose need is determined by the “parametric space” of operation; this is outer space where the speed of light is too slow to sustain immediate connectivity between objects of “astronomical” distances (Cerf, 2024).

LunaNet branches out from this developmental work. NASA’s Katherine Schauer and Danny Baird (2021) explain:

The core network framework of LunaNet is Delay/Disruption Tolerant Networking (DTN), which ensures data flows seamlessly through the network and reaches its final destination despite potential signal disruptions. In the event of a disruption between two LunaNet nodes, DTN enables the nodes to store data until the path becomes clear.

As an architecture for communication and navigation for lunar missions, LunaNet will provide “networking services, position, navigation and timing services, and science utilization services” (Israel et al., 2020: 10).

David Israel—NASA’s Exploration and Space Communications Projects Division architect and a self-proclaimed “space internetworking guy” (Olson, 2021)—and colleagues envision that LunaNet will be the backbone for communications in different lunar operations and can be used by a wide range of users, “human and robotic,” who “will experience network functionality similar to that experienced on Earth” (Israel et al., 2020: 10). The network will be built by NASA in collaboration with other space agencies, governmental organizations, universities, and the private sector. The goal is to “maximize international interoperability,” occurring at the level of standards and protocols (Israel et al., 2020: 3).

Interoperability, as defined by the Institute of Electrical and Electronics Engineers, is simply “the ability of two or more systems to exchange information and understand the information that is exchanged” (quoted in Hodapp and Hanelt, 2022: 407). This definition of interoperability is adapted and modified by different fields of study. Platform studies scholars Daniel Hodapp and André Hanelt are critical of its usefulness as an umbrella term, they note: “prior theoretical and empirical research pertaining to the interoperability concept is limited, dated, splintered into multiple research streams” (2022: 408). They maintain that different fields such as economics, information sciences, and organizational and management science use the concept on their own “with a certain degree of conceptual ambiguity” (2022: 408). Hodapp and Hanelt’s (2022: 414) targets of criticism are computer sciences and information studies, on one hand (and how this scholarship seems to reduce interoperability into technical elements), and management studies, on the other hand, where the technical elements of interoperability are dismissed in favor of institutional dynamics. For Hodapp and Hanelt (2022: 414–416) interoperability needs to be considered as a “socio-technological” concept, it is about the influence of technical standards and protocols, as well as organizational dynamics.

The point is that simple abstraction of interoperability is not a sufficient starting point for analysis, when interoperable information-exchange systems structure the world and empower the dynamics of distribution of wealth and resources in complex ways. What is needed instead is an understanding of the concrete processes where interoperability is utilized and where it functions in different ways. Thus, studies of governance (DeNardis, 2012; DeNardis, 2023), digital innovation (Hodapp and Hanelt, 2022), digitization (Thylstrup, 2019), and platforms (Bechmann, 2013) work by situating interoperability into specific contexts. Following this approach, in the next passages we will see how LunaNet’s interoperability discourse highlights both social and technical aspects. The network is designed for interoperability not only at a technical level but also with a view on how diverse service providers, both scientific and commercial could work together.

Technical diagram

How the Internet works here on Earth provides a specific lens and a point of comparison for the discussions that characterize the LunaNet’s interoperability. “Networked communications have transformed our lives on Earth” begins Israel et al.’s (2020: 1) article that aims to give a general description of the LunaNet. “We proceed with our daily lives secure in the knowledge that as long as we have a network connection, we are able to communicate with anybody else on the network simply and reliably” (Israel et al., 2020: 1). A network that effortlessly connects heterogeneous elements together, has of course done much more than brought people together. It has also “disrupted traditional business and government operations models, and created new economic markets” (Israel et al., 2020: 1). For example, “Application platforms for exchanging goods and services, as well as sharing or renting access to resources with high capital costs are all fundamentally enabled by networked communications,” Israel et al. (2020) note (p. 1). Indeed, the LunaNet is NASA’s vision to establish an infrastructure that brings a similar “service paradigm” to outer space (Israel et al., 2020: 1).

How do you build a network that connects the Earth and Moon into a similar service paradigm? The answer is node by node—the network does not need to be complete. “LunaNet will start with a simple architecture of a few nodes to meet the needs of the early missions,” Esper (2021) writes and continues that it will “evolve to meet the growing needs of a sustained lunar presence. All relay network services are not expected to be met by a single spacecraft platform, or node.”

What Esper describes is how we have come to envision networks here on Earth after-Internet. “The Internet is a meta-network of interconnected networks that communicate with a common protocol suite—Transmission Control Protocol/Internet Protocol (TCP/IP)—designed to connect any two networks,” Britt Paris (2018) writes reminding us of how the network was technically conceived through the vision of Cerf and his colleagues. Think of Internet as a network of networks. Its mode of connectivity is distributed and decentralized, it extends around the globe by connecting heterogenous networks together, provides wide coverage and even constant connectivity to multiple users when needed. At the heart of the operation is interoperability; a principle of relation which implies that networks can work together and form wholes that are bigger than their parts.

When describing LunaNet, Israel (NASA, 2021) outlines a similar conception of (inter)networking but in outer space:

once you have your LunaNet connection, then you have the ability to communicate with anybody else that’s on the network. And it’s not so much about a single connection. It’s: you connect to the network, which then gives you that interconnectivity between all the different entities.

These entities that connect can be on the Moon, Earth, or perhaps located on a space station; the network brings them together. LunaNet is designed as one shared network for all lunar missions. The key here is the networks extensibility: Israel et al. (2020) anticipate that LunaNet, as a mode of networked communications, is “ultimately contributing to the buildup of the solar system Internet” (p. 4).

While LunaNet sounds a lot like the Internet here on Earth, its operational environment, outer space, dictates a central difference between the two network models. Israel says (quoted in Olson, 2021):

The way internetworking on Earth will not work for all the different scenarios we have throughout the solar system. Earth’s internet assumes that the two ends of a “conversation” or data exchange have a fully connected, low delay path between them.

To be more exact, it is the TCP/IP that is designed for continuous end to end connectivity between heterogenous networks and devices here on Earth, and it is the TCP/IP based Internet connectivity that does not work in outer space. Israel (quoted in Olson, 2021) explains: “Our scenarios include situations such as connections light-minutes or more away and relays that may be able to see a rover but not have a simultaneous link to Earth” (Israel quoted in Olson, 2021). The long distances information needs to travel from end to end in outer space cause delays. Israel explains that their goal is “to allow internetworking in all scenarios” (Olson, 2021). What, is implemented for the LunaNet is a protocol called Bundle and it is specifically designed for outer space information-exchange needs: to achieve a connection between the networks nodes in situations of high latency (Israel et al., 2020: 3). The Bundle Protocol is built delay-tolerant by default and, as such, does not require that the ends (Earth nodes, orbital nodes, surface nodes) are fully connected. We will return to this technical difference between the LunaNet and the Internet later.

Social system

In his keynote address at the Lunar Interoperability Forum, Cerf (2024) observes, “for many people who imagine that commercialization [of outer space] is somehow decades away, several of you sitting in this room know that it is more like two years away.” For Cerf and his assumed audience, Earth’s sphere of influence is changing and extending beyond planetary borders. Cerf (2024) continues:

This is a really exciting time—a kind of renewal of our entry into space, but this time with commercial aspirations. So the support for that should come from both governments, private sector, as well as academia. And we believe that the network designed, which we worked on since 1998, can serve that particular objective.

While Cerf speaks again of the Solar System Internet, LunaNet is part of the same discourse network. Here, networked interoperability begins to reveal itself not only as a technical invention, but also as a social system.

The LunaNet is a social network long before there are actual people on the Moon using it—it is embedded in cultural and economic contexts where it is designed and used. LunaNet belongs to an era of organizational interoperability. “LunaNet is intended to allow many lunar mission users to engage the services of diverse commercial and government service providers in an open and evolvable architecture,” NASA (2025: 8) writes and later continues “Many nations, agencies, and private companies and institutions can contribute to and participate in the establishment and operation of LunaNet-compatible services.” These service providers are envisioned interoperable (NASA, 2025: 9).

Public and private entities are seen as the creators of this network (NASA, 2025: 9; Oust, 2021). As Andy Petro (quoted in Oust, 2021), the lead for lunar communications and navigation implementation at NASA, notes:

We’re looking at doing something that NASA would not necessarily build and operate, but through either commercial public private partnerships or service contract arrangements, quite possibly from multiple providers.

Kendall Mauldin, the head of NASA Goddard Space Flight Center’s Technology Enterprise and Mission Pathfinder Office, explains (NASA, 2021):

So, the vision with LunaNet is that it is an open, scalable architecture, which means that we really are counting on a very collaborative interaction with not just NASA but international and commercial entities to really help build out the infrastructure of LunaNet. So, it’s actually a really wonderful opportunity for industry—for example—to play a very strong role in the buildout of the LunaNet nodes, relays, even user services as part of the larger architecture.

Here, the discussion on LunaNet’s interoperability frames it through the lens of political willingness and technical capability to open the space sector to commercial entities. It also means that a particular ecosystem is born from these powers. In an interview with BBC, Thomas Zurbuchen, NASA’s Associate Administrator for Science, acknowledges that the Artemis mission “is not about flags and footprints [. . .]. It’s about building a sustainable presence on the Moon and to set in place the foundations for a moonbase and future missions to Mars” (Nelson, 2023). Sustainable presence on the Moon is premised on the new paradigm where governmental space agencies and the private sector operate together and, one might add, are connected to the same network.

While Cerf (2024) is one of the strongest public proponents of commercialization of this network technology, he also anticipates that commercial interests’ expansion to outer space will cause frictions. In his keynote address (Cerf, 2024), he reminds us of the 1967 Outer Space Treaty which prevents nation-states from owning land in outer space and argues that with new technologies entering that space the desires to claim territory will reemerge: “At some point, somebody’s going to say, ‘how do I register my mine or my property on the Moon?’”

The desires to commercialize space evoke colonial critiques of space exploration (see Gorman, 2025) into which the critiques of interoperability give us specific vantage point for they direct the attention to questions of networked power in a socio-technical sense. For example, Andersson Schwarz has suggested a stack model to understand how the contemporary mega platforms such as Alphabet and Facebook (now Meta) have gained their power through the “free, open, and scalable Internet infrastructure” (2017: 386). In the middle of the stack between the individual platforms and platform superstructures is the interoperable level, which creates interrelations among systems and results in ecologies of power dominated by a handful of actors (Andersson Schwarz, 2017: 379). Interoperability in this schema aggregates and accumulates power, enabling individual and local platforms to form interconnected ecosystems.

Andersson Schwarz and others, like Vili Lehdonvirta (2022: 225), work to deconstruct a very strong myth that surrounds our network culture, that interoperability at the protocol level ensures freedom and openness. To an extent, as Galloway (2004: 8) maintains, it is the opposite: “protocol is how technological control exists after decentralization.” Control does not cease to exist but takes different forms, and eventually in fact, centralized powers find a new position in the stack. Writes Elettra Bietti (2023: 4): “Moving from an Internet of networks to an Internet of platforms entailed a move from a decentralized environment where freedom seemed to be the norm, to a centralized and enclosed economy controlled by a few private actors, where freedom is more precarious.”

The early Internet’s interoperability is not abandoned but rather it is reappropriated by the platforms. Scholars like David Nieborg and Anne Helmond (2018) illustrate how interoperability allows platform ecosystems to grow. They note (Nieborg and Helmond, 2018: 203), “interoperability—the ability of software to programmatically communicate and exchange data” allows platforms to connect “with competing platforms and third-party web services and apps.” However, interoperability does not mean reciprocal relationships among different platforms but highlights asymmetries of power between those that are forced to be interoperable and those that can provide its infrastructure (Nieborg and Helmond, 2018: 203; Cf. Burgess and Baym, 2020: 24).

Interoperability in our network culture is power; it boosts communication and exchange of data and provides the platform owners with the means to control and police data flows (Halavais, 2019). The interconnected (or interoperable) global market has enabled platform companies to find regulatory loopholes that best serve their platform businesses, as well as to operate beyond nation-states (Li, 2024). According to Lehdonvirta (2022: 213), platforms “are states without estates, empires in the cloud” and “constructing entirely new markets not delimited by territory.” Again, the unifying force within this model is interoperability, the capacity of disparate elements to operate in concert. In interoperability, the elements do not remain isolated; instead, they function in unison, constructing a new whole whose logic and structure can differ from the territorial and institutional frameworks.

While media theorists like Nick Couldry (2025: 162) see imposed interoperability across all platforms as a pathway to more “robust communication ecology and a better functioning market,” others are more skeptical. Interoperability harmonizes, but with brute force, Nanna Bonde Thylstrup (2019: 66) for example maintains, noting that its harmonizing power is one of compression. Popiel and Vasudevan (2024: 4) explain such power through the operation of scaling—extending the systems’ influence and homogenizing systems, policies, and services along its way. If we examine the terrestrial Internet, we observe that opening the network to commercial players helped expand and democratize the network in the short term. However, in the long term, it seems that power has been compressed in the hands of the few who now harmonize or homogenize network logic to their own ends (see Popiel and Vasudevan, 2024).

Thus, the end of this process of compression is what both Bechmann (2013: 19) and Sutor (2011: 214) call intraoperability. Intraoperability speaks of the uneven power relations where software providers are dominant and want to “suck all-important data and processing into the central software ecosystem” (Sutor, 2011: 214). While physically, the networks expand through interoperability, socio-technically (and culturally, politically, and economically), the world becomes smaller and compressed. In terms of interoperability, the distinction between here (Earth) and other locations in the solar system (there) ceases. Instead, everything is part of the same whole, unified by the network that connects everything within the same system.

There

What if there is a difference? According to anthropologist Valerie Olson (2023: 31), “Spaces outside Earth’s atmosphere are, for unaided humans, unsurvivable. And yet all societies include those outer spaces—and their things and entities—within the bounds of shared experience.” She continues, “They are apart but not disconnected, here but not here” (Olson. 2023: 31). Drawing on the sociology of space, Olson (2023: 36) explains how since the 1950s, many nation-states and space agencies desire to form a closed system at the scale of the solar system and thus for her, space missions that are redrawing the relational boundaries for our existence and experience are extensions of economic and ecological power. However, she also notes that they can potentially give rise to new ways of thinking and being and belonging in outer space (Olson, 2023: 36–37). Olson’s (2023) “more-than-terran” thinking allows us to begin forming an argument that there (outer space) is not the same as here (on Earth), and it does not need to be.

To be clear, my distinction between here and there is as much philosophical as it is political. I draw on Stengers (2011a: 64), who insist that the contrast between here and there should not be stated in terms of distance or space. Here and there are not interchangeable locations but situated events. To say that here and there are the same would be to abstract them from their uniqueness. Thus, arguing that LunaNet is like the Internet and that the Moon, Mars, and other celestial bodies can be similar to Earth is not a statement of fact but a sleight of hand that must be made visible because, as shown already, it is a pathway to a particular form of power.

Technical diagram

Britt Paris (2021) argues that time and temporality are central design constructs giving networks and our experiences of networking their shape. In the case of the LunaNet, we see two configurations of temporality: networked services that operate in real time and those that operate with delay (NASA, 2025: 21). These two configurations are expressed in the “LunaNet Interoperability Specification Document,” (NASA, 2025) whose fifth version was published in 2025, which sets to “define a framework of mutually agreed-upon standards to be applied by users and service providers in a cooperative network supporting missions on and around the Moon” (Manning, 2023) It is the latter configuration, the one built around delay-tolerant networking and the aforementioned Bundle Protocol, that is of our interest here for it will allow us to think of networked relations as extrarelations, those beyond immediate or unifying connectivity.

To make this point, we need to investigate LunaNet’s technical specificity: delay-tolerant networking. “Network layer services over links which are characterized by disruption/delay or where a robust end-to-end path is not available (or at full bandwidth) will be supported by DTN [delay-tolerant networking],” NASA (2025: 21) explains and specifies that Bundle Protocol will be used in these occasions. Indeed, the Bundle Protocol is fundamental to delay-tolerant networking (Israel et al., 2020: 3–4). The protocol is specifically designed for outer space, enabling networking that can sustain delays caused by for example the physical distance between two ends.

In the LunaNet’s case, this is the delay between the Earth and the Moon. Israel et al. (2020) outline how this mode of networking works and how the Bundle Protocol enables the transport of “self-contained data units” known as “bundles.” The bundles are equipped with metadata which include information about its transmission: where the bundle comes form, where it is going, and how urgent it is. If a connection between two nodes of the network is established, a new node takes custody of the bundle; if no connection is immediately available, the bundle is stored and forwarded when an opportunity to transmit presents itself (see also Barkley et al., 2010). While TCP/IP does not tolerate delays beyond a few seconds, the Bundle Protocol allows bundles to wait for hours, days, or longer for a connection to appear.

Because of the way the Bundle Protocol enables connectivity in the form that is not continuous, LunaNet is said to be a “flexible and extensible” infrastructure beyond planetary borders (Israel et al., 2020). Therefore, LunaNet operates in ways where real-time connectivity is not a default condition, but possibility among a range of others. In practice, in this model the bundles are sent from one hop to another—for example, from Earth to one satellite to another—until each bundle reaches its end (Scott and Burleigh, 2007). When store-and-forward nodes are unavailable, and bundles are put on hold until a connection appears, delay-tolerant networking turns the attention from the experience of unity, totality, or continuity to forms of connectivity that are sporadic, fragmented, or episodic. In this diagram connections take the form of extrarelations, which work across discontinuities in connectivity, can survive gaps within an infrastructural constellation, because they do not need to a continuous whole to maintain the network and its identity. In other words, within this framework, extraoperability describes the capacity of two or more systems to exchange and interpret information without being integrated into a single, unified structure or a whole.

Social system

“LunaNet is the first step in creating the Solar System Internet and will keep explorers connected to Earth,” a NASA video which featured the two astronauts mentioned in the beginning proclaims (NASA’s Goddard Space Flight Center, 2021). “The priority for our journey to the Moon is discovery—learning how to live in space so that one day we can establish a human presence on Mars,” the video (NASA’s Goddard Space Flight Center, 2021) continues, framing the network as a lifeline for astronauts who are no longer physically on their home planet. The video begins with the artist’s rendering of two astronauts near the Shoemaker crater, as described earlier. The implication of the image and the video is simple. To be is to be connected; to be online is to be alive in outer space.

Imagine our two astronauts, one plugging in the cable, and the other monitoring the tablet, but now the surface on which they stand has a reddish hue for they are on Mars not on the Moon. Now imagine that their lander has a leak in its oxygen production system and there is oxygen left for ten minutes if the generator does not restart. This a problem that has a simple solution, but to solve it the astronauts need to contact Earth. Their spacecraft has a radio link to Earth, but connectivity is sporadic. “If there is a problem and the spacecraft tells us, we won’t know for 13 minutes, and then even if we react straight away it’ll be another 13 minutes before our instructions get back to Mars—there’s a lot that can happen in half an hour at Mars,” describes Thomas Ormston (2012) on a blog post that deals with the time delay between the two planets. Connection with Earth comes with a communication delay between the planets that no existing network technology can solve. In emergencies like this, delay-tolerant networking cannot save the astronaut. It ensures eventual delivery of data but not immediacy of communication with Earth which would be required. In other words, extraoperability creates a functional communication transmission system between here and there, but the way it connects does not erase the difference (spatial and temporal) between its parts.

In a Wired interview from 2013, Cerf says that with this protocol, real-time connectivity is lost and we must adapt:

We will need to learn how to go from what appears to be an interactive real-time chat, like one over the phone, to probably an email-like exchange, where you might have voice and video attached but it’s not immediately interactive (Mann, 2013). ²

Since delay is inevitable, LunaNet establishes connection, but in doing so, it does not necessarily unify time or space. In this way, LunaNet’s temporal design is not relatable to the common experience of the Internet: it does not create “immediacy” (Parks, 2005: 23), nor will it reorganize space in terms of “continuous flow” (Starosielski, 2015: 17). LunaNet in practice shows us that connectivity, like the unity it is supposed to bring, is a construct of mediation and not something that is given. Kittler (2017: 14) of course argues that theories that try to differentiate “between delay and simultaneity, are myths,” implying that our experience of real time is already constructed by fast-enough processing that that can make delay imperceptible to our senses.

Nevertheless, the concept of extraoperability invites us to imagine networking as a process that rather than merely unifying its parts under the logic of extensibility and scalability generates something new instead. For example, since very early on in the development of the Bundle Protocol and delay-tolerant networking some of the makers have discussed about the possibility of Earth, the Moon, and Mars each having their own network domains, with information transmitted from the terrestrial Internet to the Martian Internet and so on (Hooke, 2001; See also Burleigh et al., 2003).

More recently, in 2023, Interplanetary Networking Special Interest Group (IPNSIG) of the Internet Society launched a report (Internet Society Interplanetary Chapter, 2023) that arrived at a similar conclusion. It suggests separating the network between the Earth domain and the Moon domain. The reason is that the Moon’s craters and underground tunnels are “not directly accessible to communications from Earth” and would thus become subregions of the Moon domain only (Internet Society Interplanetary Chapter, 2023: 42). More regions and passageways connecting them are expected to follow as the size of the network grows (Internet Society Interplanetary Chapter, 2023: 66). Mars, the Moon, and Earth remain distinct regions even in connection. Each with their own temporal and spatial conditions, and indeed in the future each with their own Internets. This recalls Parisi’s (2013: 91) reading of Whitehead’s “mereotopological schema of finding the conditions for novelty” where a whole is “neither smaller nor bigger than its parts but is split into parts or partialities that do not necessarily communicate with one another.”

To rephrase, while the discourses of interoperability strive to explain how unifying principles like the “common good” require that heterogeneous components bear similarities at the level of design, data sharing, and information processing so that we can seamlessly switch between systems (Couldry, 2025: 162), it appears that extraoperability by default prioritizes constructive independence and separation instead. LunaNet’s delay-tolerant networking and how it empowers the functioning of its independent parts without a continuous whole and how the Bundle Protocol allows the nodes to take custody of information even in the absence of an immediate connection to forward it could be seen as an example of the latter. It does not require or predispose continuity but rather composes the network of discontinuous parts, of regions that are autonomous. If this matters politically, it is because discontinuous connectivity marked by delay provides alternatives to processes that deliberately blur the boundaries between spaces for the sake of neoliberal governmentality (Parisi, 2013: 160). It refutes the fantasies where interoperable interplanetary network is envisioned to erase “the last remaining geographical distance,” annihilate “the last void of temporal delay,” and create a real-time “imperial space” (Damjanov, 2015: 903).

LunaNet’s discontinuous connectivity shows that we do not need to think of a network as an ontological whole, a protomodel that compresses discrete regions into a unified totality. This is important because in outer space a continuously connecting whole is a physical impossibility (at least for now), but also because we are now forced to think what in fact are the continuities that are brought along in connection (cf. Stengers, 2011a: 197).

“Each network defines its own power relationships”, Castells (2011: 775) maintains and later continues that “forms of power follow the logic of network-making” (p. 776). Extraoperability, for me is an attempt to formulate a logic of network power around Whitehead’s metaphysical claim: “There is a becoming of continuity, but no continuity of becoming” (Quoted in Stengers, 2011a: 197, italics removed). In extraoperability, relations become determined by parts rather than wholes, discontinuities favored over continuities, power is distributed and localized rather than centralized, and geocentric perspectives give space to interplanetary multiplicities. As such, extraoperability is at the same time a technical proposition to think of the many instead of one, and a social and political proposition to examine how networks can decentralize and recentralize domains of power, but it also requires that we abandon the natural (or naturalized) centers of experience like our beloved Blue Planet. The distinction between here (Earth) and there (outer space) is important because outer space is extraterrestrial and so will be the networks that operate there. Or to rephrase, herein lie the political stakes presented in this article: a move to outer space does not necessarily mean the continuity of everything we have here. An analysis of the LunaNet shows instead that networks and, by extension, political models, modes of social organization, even economic principles could be reenvisioned there, in outer space.

A conclusion

If we believe in Kittler’s (1999, 2006) claim that media organize our relations in the world, the LunaNet discourse signposts two directions how this happens. The first is the Heideggerian pathway where media technologies themselves matter less than their capability of reducing the world into a resource (Heidegger, 1977; see also Parisi, 2019; Zechner, 2018). Here we can accept LunaNet as just one more network that grows on top of the Internet’s existing standards (Cf. Hu, 2014: 7–8) and eventually contributes to its extension and extraterrestrial upscaling. LunaNet thus can be seen as a variation of the same old forms of power; the empires that are now in the cloud are moving to the solar system next.

LunaNet is going to be interoperable with other networks, such as the Internet, and will allow connections across various types of networks. This is not the problem. The pursuits for social interoperability, that the network design makes possible that different stakeholders use and develop it, is also not the problem, not exactly. Rather, the problem is seeing interoperability as a solution that allows us to simply plug heterogeneous extraterrestrial spaces into our existing ecosystems—with the associated implication that interoperability inherently is the networked condition of freedom, openness, common good. We already know that the terrestrial Internet neither was nor became a “’privilege free’ [. . .] space of freedom, escape, and libertarian self-interest” (Chun, 2021: 9). Equally, I think we know that projects of extension and scaling, the ones that aim to create a whole, easily become projects of destruction and erasure of differences.

Some voices within NASA warn us that the aspirations for LunaNet’s commercialization enabled by technological interoperability could empower the privileged few. One of those sounding the alarm is Israel, who says, “We need to make sure that this [LunaNet] doesn’t become one commercial provider with proprietary things where everybody that’s going to the Moon has to buy this company’s systems” (quoted in Moss, 2021). This concern is reminiscent of the critical platform scholars who show that while interoperability allows a network’s extensibility, it also compresses what it extends (Bechmann, 2013; Sutor, 2011; Thylstrup, 2019). In other words, interoperability does not necessarily expand the world but contracts it by harmonizing everything under a single framework. If that framework works for you, great. Otherwise, little can be done. Interoperability narrows possibilities of all-new network models; it also flattens our world: everything that is, is here and now.

There is a second path, but following Kittler, we need to turn Heidegger’s idea upside down and see how thinking could be technologically reconfigured (Parisi, 2019; Zechner, 2018: 130–131). ³ To rephrase, since networked relations do not emerge naturally but require mediation, it makes sense to try to develop new thinking from and with the technologies. I have suggested turning the attention to what makes LunaNet unique: the delay-tolerant networking and how it provides us with the framework of extraoperability to contextualize the discontinuous networked relations between the Earth, the Moon, and Mars. In other words, extraoperability is a way of thinking about networked relations beyond the idea of unification or constitution of dominant sociocultural or economic wholes. Here, the notion of extraoperability is, of course, an academic provocation, a frame of thinking, rather than a fully fleshed-out concept that space agencies or similar actors could adapt.

At least in theory, extraoperability allows us to oppose the claims of the builders of the interplanetary network, who insistently invoke the Internet as LunaNet’s map and model. The Internet as a point of reference is no surprise: the terrestrial Internet has become the discourse network of our time, one that explains how our culture is based on specific practices of handling information (Kittler, 2022: 369). However, LunaNet promises to be something else than the Internet because it not only enables networking of celestial objects for the first time but also configures connectivity in a form that is incompatible with many continuities.

Extraoperability can help us to think of networks in which the whole does not compress its parts but empower them to form new openings. For example, Israel is pluralist in his pragmatism: “The more LunaNet Service Providers and technology providers the better” (quoted in Moss, 2021). Perhaps we should not take this statement as merely a stance on economic liberalism in favor of competition against monopolies or central planning but as an ontological proposition for diversity, which can go beyond global (or interplanetary) capitalism.

This aligns with what Stengers (2011b: 371–372) foresees as our future lifeline, not only in outer space, but also on this planet. For her, politics does not mean tolerating differences in favor of a unified subject but entails building modes where the “copresence of practices,” sometimes problematic, become actualized (Stengers, 2011b: 372). What the rise of LunaNet, together with the Artemis program, actualizes is a new situation—not characterized by the closure of a whole (the Moon as a continuation of humanity’s presence), but by the creation of new parts that are both related and separated by the network: a cut where we are split between different lifeworlds, Earth, the Moon, and what comes next. This is indeed a situation in which coordinates become determined by media (Kittler, 1999: xxxix) and a situation of exchange where parts will “undergo a transformation” (Stengers, 2011b: 372) in connection. The LunaNet as an extraoperable network reorganizes the world not only because of what it connects, but because of how it connects.