The Seductive, and Deceptive, Promise of On-Orbit Servicing
For more than sixty years, the space industry has operated on a simple, brutal, and expensive philosophy: “one and done.” A satellite – often the product of a decade of work and costing hundreds of millions, or even billions, of dollars – is built with extreme care. It’s launched on a rocket, a controlled explosion that represents one of the most violent events a piece of hardware can endure. If it survives the launch and successfully deploys its solar panels and antennas, it begins its mission. From that moment, the clock is ticking.
If a single, non-essential component fails, the satellite’s capability is degraded. If a major component fails, or if it simply runs out of the tiny amount of propellant it uses for “station-keeping” to maintain its orbit, the mission is over. The billion-dollar asset becomes just another piece of junk, a silent ghost in the void.
On-orbit servicing (OOS), is presented as the elegant, logical, and revolutionary alternative to this paradigm. It’s an idea that is intuitively appealing because it mirrors every other complex industry on Earth. Proponents frame it as the “new frontier” of space services, the obvious next step in a mature space economy. They ask a simple question: why would we throw away a billion-dollar asset when all it needs is a simple fix or a tank of gas? We wouldn’t do this with a jet, a supertanker, or a car. Why do we do it in space?
This vision is one of a bustling, sustainable, and resilient orbital ecosystem. It’s a future with “roadside assistance” for spacecraft. In this vision, a fleet of sophisticated robotic servicing vehicles would patrol the most valuable orbital highways, primarily in Low Earth Orbit (LEO) and Geostationary Orbit (GEO).
These servicers would perform a whole catalog of capabilities. This menu of services is ambitious and broad, covering almost every phase of a satellite’s life:
- Refueling: A robotic servicer would autonomously rendezvous with a client satellite, connect to a fuel port, and replenish its propellant, extending its operational life by years.
- Repair: This is the most complex vision, where a servicer’s dexterous robotic arms could fix or replace malfunctioned components, such as a faulty electronics box, or assist in a failed deployment, like a solar array that got stuck.
- Relocation: A servicer could act as a “space tug,” moving a satellite to a new orbit to meet changing market demands. This also includes correcting orbital insertion errors – saving a satellite that was delivered to the wrong altitude by its rocket – or towing a defunct satellite to a “graveyard orbit” at its end of life.
- Upgrades: Perhaps the most futuristic capability, this involves attaching new, modular payloads to an existing satellite. A servicer could fly up, remove an old sensor or computer, and bolt on a new, more powerful one, theoretically keeping a 20-year-old satellite “current” with modern technology.
- Inspection: In this scenario, a servicer would perform a close-up visual analysis of a malfunctioning satellite to help engineers on Earth diagnose a problem, much like a mechanic looking under the hood.
- Active Debris Removal (ADR): This function is often bundled with the servicing business case. The same robotic technology used to grab a client satellite could be used to grab a piece of “space junk” – a derelict rocket body or a dead satellite – and actively push it into the atmosphere to burn up.
The stated benefits of this vision are, on their face, transformative. Government agencies and private ventures claim this new paradigm will make spaceflight more affordable by maximizing the life of assets. They claim it will make space more sustainable by actively cleaning up orbital debris. And they claim it will make space architectures more resilient, offering a way to repair and maintain critical national security and scientific satellites, like the International Space Station or the Hubble Space Telescope, which required human-crewed servicing missions.
This is the seductive promise of on-orbit servicing. It’s a compelling narrative of progress, logic, and sustainability.
It is also, upon critical examination, a mirage. The entire pro-servicing argument is built on a foundation of flawed analogies, broken economics, and a willful disregard for the extreme, non-terrestrial realities of the space environment. The proponents of servicing are selling a 1990s-era solution to a 2020s-era market that has already moved on.
The seductive analogy of “servicing a car” is the first and most deceptive flaw. A car is designed to be serviced. It operates in a dense atmosphere, is bound by friction and gravity, and is easily accessible in a one-g environment. A satellite is a delicate, bespoke object, often wrapped in fragile gold-foil insulation. It moves at over 17,000 miles per hour in a vacuum, where there is no friction and temperatures swing by hundreds of degrees. An on-orbit “repair” is not a simple task. It’s a multi-billion dollar robotic mission of such extreme complexity that it has only been done successfully on one object (Hubble) and required the presence of human astronauts.
This is complicated by a deliberate vagueness in the industry’s own marketing. The term “in-orbit servicing” is a “big tent,” intentionally bundling simple, achievable missions with highly complex, speculative ones. For instance, the industry points to Northrop Grumman’s Mission Extension Vehicle (MEV) as proof that servicing “works.” But the MEV is, in reality, a clever “piggyback” tug. It docks with an old satellite and then uses its own thrusters and fuel to act as a new engine for the combined-spacecraft stack. It does not refuel the client. It does not repair the client. It does not upgrade the client. This is a important distinction. The success of a relatively simple “tug” service says nothing about the economic or technical viability of the far more complex and risky robotic “repair” or “refueling” missions that form the core of the speculative business case.
This bundling is a bait-and-switch. It allows the industry to claim victories on simple tasks while the core premise – that complex, autonomous robotic surgery in space is a viable business – remains entirely unproven. The case against in-orbit servicing isn’t just one argument; it’s that every pillar of the pro-servicing case – economic, technical, environmental, and geopolitical – collapses under the weight of reality.
The Flawed Economic Premise
The most compelling argument for any new technology is that it saves money or creates new, immense value. The advocates for in-orbit servicing insist it will do both, unlocking decades of “bonus” revenue from satellites that would otherwise be decommissioned. But a objective analysis of the business case reveals that it’s not just weak; it’s fundamentally broken. The entire model is a high-cost, high-risk solution competing in a market that is rapidly and relentlessly moving toward low-cost, disposable assets.
The Servicing vs. Replacement Cost Fallacy
The core economic argument for servicing rests on a simple, and increasingly false, premise: that it is cheaper to fix a satellite than to replace it.
Market research conducted with commercial satellite operators – the supposed customers for this new industry – has established a very clear and unforgiving benchmark. This “50% rule” states that servicing is only considered economically viable if the total cost of the servicing mission is less than 50% of the cost of launching a replacement satellite.
This rule is the first rock upon which the servicing business case shatters. A servicing vehicle is not a simple “tow truck.” It is one of the most complex robotic spacecraft ever conceived. It must be designed with extreme reliability, outfitted with multiple, high-resolution sensors, and equipped with some of the most advanced, dexterous robotic arms ever built. It must then be launched into a precise orbit. The development cost for such a vehicle can run into the hundreds of millions or even billions of dollars. The launch itself, a single rocket, adds tens of millions more to the price tag.
To have any hope of a positive return on investment, a single servicer would need to perform multiple missions, hoping from one client satellite to the next. One analysis showed that a theoretical refueler would need to successfully service three to five customer satellites just to break even on its own development and launch cost.
This math simply doesn’t work in a high-risk environment. What happens if the servicer’s first mission is a success, but it collides with the client on its second attempt? The entire multi-billion dollar business model is wiped out in an instant.
This already-shaky model is aimed at a customer base that is, itself, shrinking. The primary market for this kind of high-value servicing was traditionally the fleet of large, expensive Geostationary Orbit (GEO) communications satellites. These are the bus-sized spacecraft that linger in a fixed spot in the sky, beaming television and data to entire continents. But this market is no longer the titan it once was. Orders for new GEO satellites have dropped significantly in recent years. Operators are now assessing life-extension on a “case-by-case” basis, not as a guaranteed new revenue stream. The expected “gold rush” of customers is, in reality, a small and wary group of operators who are far more interested in the next generation of satellites than in paying to keep their old ones alive.
The business model for servicing is, in effect, financially inverted. It places 100% of the development cost and all of the initial risk onto the servicing provider (often a startup or a new venture). The customer (the large satellite operator) remains risk-averse and has shown no willingness to pay for unproven technology.
This creates a fatal capital-investment gap. Commercial satellite operators have been explicit: they do not want to pay for the development of new technology. Investors, in turn, are not willing to pay the cost above a nominal, proven price. The servicing provider is stuck. It must first raise and spend billions of dollars to build the “gas station” (the servicer), launch it into orbit, and prove that it works, all with no guaranteed customers. The satellite operators, meanwhile, are waiting for the service to be as cheap, reliable, and routine as a terrestrial delivery. This is a classic, and in this case, insurmountable, capital hurdle. The market has extremely high entry costs for an unproven service, and the very customers who supposedly need it are unwilling to help fund its creation.
The “Chicken and Egg” Market Failure
Even if the cost model wasn’t prohibitive, the market for servicing is stuck in a logistical paradox that has, for decades, crippled its development.
This is the “chicken and egg problem,” an issue identified by government agencies and market analysts alike. The paradox is simple and, so far, unsolvable:
- The “Egg”: Satellite operators are hesitant – to the point of refusal – to design and deploy satellites that are serviceable. This would mean adding extra mass for a standardized grapple fixture, designing a universal “fuel port,” and ensuring electronics are modular. They won’t make this investment until servicing is a fully mature, commercially available, and, most importantly, cheap commodity.
- The “Chicken”: Servicing providers are hesitant – to the point of impossibility – to develop and launch their billion-dollar robotic servicers until there is a clear user base of serviceable satellites to work on.
This isn’t just a market-timing issue; it’s a deep-seated technical problem. For servicing to be economical and scalable, there must be industry-wide standards. Proponents have long envisioned a “USB port” for satellites – a standardized, universal interface for refueling, data transfer, and power.
This “USB port” does not exist. It has never existed. And it shows no sign of emerging.
Every satellite from every manufacturer is a bespoke, custom-designed piece of hardware. A servicer designed to refuel a satellite built by Thales Alenia in Europe would need a different nozzle, grapple system, and software suite than one designed for a satellite built by Maxar in the United States. This lack of standardization means that every single servicing mission becomes a new, custom-designed, eye-wateringly expensive research and development project. It completely destroys any hope of building a scalable, low-cost operation.
A deeper analysis suggests this “market failure” is not an accident. It is a permanent structural barrier, actively (if passively) maintained by the incumbent satellite manufacturers whose entire business model is directly threatened by the concept of servicing.
The “chicken and egg” problem is not an unfortunate friction in the market. It is a rational, logical, and profitable outcome for the companies that dominate the space industry. The very manufacturers whose revenue comes from selling new, “one and done” replacement satellites every 10 to 15 years are the same companies that would need to agree on and implement the standardized “USB port.”
Why would they do this? Why would a manufacturer, whose profits depend on replacing satellites, invest in a technology that would allow a third party to make their products last twice as long? Doing so would, quite literally, cut their future revenue streams in half. The “failure” of the market to produce these standards is not a failure at all. It is the silent, rational, and perfectly logical outcome of a market protecting its own profits. The servicing industry isn’t just fighting physics; it’s fighting the entrenched and powerful business model of its own potential suppliers.
An Uninsurable Venture
A final nail in the economic coffin is that the risks introduced by in-orbit servicing are so novel, so extreme, and so high-stakes that they are functionally uninsurable, creating a toxic environment for private investment.
The space insurance market is already one of the most volatile financial markets in the world. It is not like car insurance, with millions of “safe” customers. It’s a small, niche market characterized by a very small pool of exceptionally high-value assets. This means a few large claims can – and regularly do – wipe out an entire year’s worth of premiums for the whole industry. In 2023, for example, claims from just a couple of failed GEO satellites exceeded the market’s total premium income by hundreds of millions of dollars.
This is not an industry that is eager to take on entirely new, untested, and un-priceable categories of risk.
Current satellite insurance focuses almost exclusively on the “launch” phase – the 30-minute ride to orbit and the first year of operations. This is a risk that underwriters, after 60 years, finally understand and can model. In-orbit servicing introduces a completely new and terrifying risk profile: a high-speed, autonomous rendezvous and a complex, hands-on robotic interaction in orbit.
Insurers cannot price this risk. The questions they face are unanswerable:
- What is the premium for a mission where the servicer’s robot arm accidentally damages or destroys the client’s multi-hundred-million-dollar satellite?
- What is the premium for a failed docking attempt that sends both the servicer and the client tumbling out of control, creating a massive debris cloud that threatens other satellites?
- What is the premium for third-party liability – where the servicing mission collides with a different satellite, one not involved in the operation, belonging to another company or another country?
The very nature of satellites – high-value, impossible to inspect, and (until now) impossible to repair – is what makes them so difficult to insure. Servicing attempts to solve the “repair” problem, but in doing so, it creates an “inspection” and “collision” risk that is infinitely worse. This un-priceable risk is a red light for investors. Without insurance, a company can’t get financing. Without financing, a company can’t build a servicer. This is the last, and perhaps most final, break in the economic chain.
The Disposable Revolution: Why New Models Obsolete Servicing
The most compelling argument against in-orbit servicing is not that it’s too expensive, too risky, or too difficult, though it is all of those things. The most compelling argument is that it is already obsolete.
The entire concept of in-orbit servicing is based on a 1990s-era model of space – a model defined by a few, extremely high-value, monolithic satellites. That era is over. A new economic and design philosophy has completely revolutionized the industry, one based on mass production, disaggregation, and rapid replenishment. This new model has rendered the very idea of “repairing” a satellite in orbit fundamentally and permanently absurd.
The Economics of Mass-Produced Constellations
The debate over servicing is, at its heart, a debate between two competing business models.
The traditional model, which servicing hopes to address, involves building a satellite the size of a school bus. It costs $500 million, takes eight years to design and build, and is launched to a single slot in Geostationary Orbit (GEO).
The new model involves mass-producing satellites the size of a mini-fridge on an automated assembly line.
The Starlink constellation provides a stark and unassailable economic case study. While exact figures are proprietary, based on industry analysis, a single Starlink Version 2 Mini satellite is estimated to cost around $1 million to build. The marginal launch cost per satellite, using a reusable Falcon 9 rocket, is estimated to be around $680,000. This brings the total cost to build and launch one highly capable satellite to approximately $1.7 million.
Now, apply the “servicing” logic to this asset. The economic premise of servicing is to save a high-value satellite. But it is economically insane to design, build, and launch a $200 million robotic servicer to perform a $50 million servicing mission to refuel a $1.7 million satellite. The cost of the service is orders of magnitude more than the cost of the asset. It would be like a city calling in a $100,000-per-hour heavy-lift helicopter to change a flat tire on a municipal bus. The math is not just bad; it’s nonsensical.
This isn’t a niche market. These new “mega-constellations” in Low Earth Orbit (LEO) are not a “future” concept; they are here, and they are actively disrupting and displacing the “traditional” GEO market that servicing was supposed to target. The very customer base that high-value servicing was designed for is evaporating, being replaced by a new model that has absolutely no need for it.
The entire pro-servicing model is predicated on the “monolithic” satellite paradigm. The new, dominant paradigm is the “disposable” or “replenishment” model. These two are in direct economic competition, and the disposable model is winning. Proponents of IOS are trying to sell an expensive, complex life-extension service to a market that has embraced a cheap, disposable, and rapid-replacement model. They are selling buggy whips in the age of the automobile.
Resilience Through Replenishment, Not Repair
Proponents of servicing argue that it offers “resilience” by allowing for the repair of failed assets. This is another example of a word co-opted from the old paradigm. The new constellation model achieves a far more robust, flexible, and anti-fragile “resilience” by design, without ever having to attempt a high-risk robotic repair.
This is the concept of “graceful failure.” A monolithic, $500 million satellite is a single point of failure. If its main processor or transmitter fails, the entire capability is lost. The system has no resilience.
A distributed constellation, by contrast, is inherently resilient. In a constellation of 2,000 satellites, the failure of one, ten, or even fifty units is a rounding error. It is a minor, expected operational hiccup. The network, a “mesh” in the sky, automatically and instantly routes data around the dead satellites. This is “graceful failure,” and it is a level of resilience a monolithic satellite can never achieve.
When a satellite in one of these large constellations fails, the operator doesn’t scramble to schedule a multi-million-dollar repair mission. It doesn’t even care. It simply launches a replacement.
This is the “rapid replenishment” strategy. Operators of large constellations like Starlink are already “continually launching satellites as part of the normal replenishment cycle” just to replace older units that are aging out. A failed satellite isn’t a crisis; it’s a notification to the logistics team to move a new, identical, mass-produced satellite to the top of the launch queue. This replenishment strategy is a faster, cheaper, and 100% more reliable way to maintain a constellation than a high-risk robotic servicing mission.
Designing for Obsolescence as a Feature
The final flaw in the servicing model is that it mistakes “obsolescence” for a problem. In the new paradigm, it’s a feature.
Proponents of servicing dream of a satellite “bus” (the main body) that lasts for 30 years, with its “payload” (the computers and sensors) being robotically upgraded every 5 years. This assumes that a complex, risky, and unproven robotic surgery is the preferable way to stay current.
The constellation model offers a far more elegant and powerful solution. LEO satellites are, by nature, short-lived. They have operational lifetimes of 3-7 years, much shorter than their GEO counterparts, because of factors like increased atmospheric drag and more frequent battery-draining eclipses. But this short lifespan is a feature, not a bug.
Shorter satellite lifetimes “facilitate more rapid introduction of new technology.”
Consider the two upgrade paths:
- The Servicing Model: After 15 years, a satellite operator pays $50 million for a robotic servicer to fly to its aging satellite and attempt a risky “surgery” to install one new sensor onto a 15-year-old “bus” that still has 15-year-old wiring, 15-year-old power systems, and 15-year-old batteries.
- The Replenishment Model: A constellation operator simply builds its next batch of 50 satellites with that new sensor already integrated at the factory. These new, next-generation satellites are then launched as part of the normal, scheduled replenishment cycle.
The replenishment model is faster, cheaper, and eliminates the 100% of the risk associated with a failed upgrade mission. It also ensures that the entire satellite – the bus, the power systems, the sensors, the computers – is brand new, not just one component. This model allows for a total-system tech refresh every 3-5 years, a pace that the old, monolithic model could never hope to match.
This fundamental difference in design philosophy is the ultimate case against servicing. The following table provides a direct comparison of these two competing and mutually exclusive business models.
Even if the economic case for servicing weren’t broken, and even if the entire business model weren’t already being rendered obsolete, the case against would still be overwhelming on technical and operational grounds alone. Proponents of in-orbit servicing, when sketching out their business plans, vastly underestimate the extreme, unforgiving difficulty and risk of autonomous robotics in space.
The Rendezvous and Proximity Nightmare
The single most dangerous and unproven part of any servicing mission is the “rendezvous and proximity operation” (RPO), or RPOD (which adds “docking”). This is the phase where the servicer autonomously approaches the client satellite, closing the final few kilometers, and then makes contact.
The mission was a disaster. DART performed as planned for the first several hours, but then the spacecraft began to suffer from “incorrect navigational data.” Its onboard sensors were not providing an accurate picture of its location relative to the target. This led to “higher-than-expected propellant consumption” as the autonomous flight computer desperately tried to correct its course based on bad information. Just 11 hours into its 24-hour mission, DART detected it was out of propellant, initiated its retirement sequence, and “ended up colliding with the communications satellite” it was intending to maneuver around.
A subsequent mishap investigation found that this wasn’t just a hardware glitch. The collision was the direct result of “inadequate guidance, navigation, and control software development processes” and, most damningly, a “poorly managed risk posture.” This failure showed that the process of building, testing, and managing these autonomous systems is fundamentally unreliable. The DART mishap is the perfect, real-world example of what happens when the immense complexity of autonomous RPO meets reality. The result is a collision.
RPO remains “particularly challenging” because it requires a level of autonomy, sensor technology, and precise maneuverability that is at the absolute bleeding edge of our current capability. A failure in this phase is not a minor setback; it’s a catastrophic, debris-creating event that destroys both the servicer and the client.
The “Non-Cooperative” Majority
The few successful RPO missions that the industry can point to, such as cargo resupply flights to the International Space Station (ISS), all involve a “cooperative” target.
A “cooperative” client is a satellite that is designed to be docked with. It is an active and willing partner in the dance. It has standardized docking ports, special grapple fixtures for a robotic arm, and, most importantly, it has two-way communication. It actively shares its precise position, velocity, and orientation with the approaching vehicle.
The satellites that servicing proponents want to fix – the ones that are “malfunctioned,” “out of fuel,” or “dead” – are, by definition, “non-cooperative.”
A “non-cooperative” client is a completely different, and terrifying, target. It has no docking ports. It has no grapple fixtures. It is not communicating. And worst of all, it is very likely “tumbling” – spinning uncontrollably on one or more axes.
This means a servicing mission to a non-cooperative target is not “docking.” It is a high-stakes robotic capture mission. The servicer must approach a silent, spinning, unpredictable object of incalculable value. It must autonomously match its rotation in three-dimensional space with picosecond-perfect timing. And then, it must extend a robotic arm to grab onto a part of the satellite – like a fuel nozzle, an antenna, or the engine bell – that was never, ever designed to be grabbed.
This is not a service; it’s a gamble. It is an order of magnitude more difficult, more dangerous, and more likely to fail than the “cooperative” docking that is already so difficult. And yet, this is the entire business case.
The Standardization Chimera
This technical hurdle – the non-cooperative target – loops directly back to the economic failure. Because every satellite is a bespoke, non-cooperative target, the lack of standards creates an impossible technical, and therefore business, hurdle.
This lack of standardization – the non-existent “USB port” – means a servicer cannot be a “one size fits all” tool. A scalable business requires a standardized, repeatable service. But because every satellite is different, the servicing industry is not a “gas station” business; it’s a “custom, one-off robotic R&D” business.
A servicer would need to be a “Swiss Army Knife” with a different tool for every job, a technical impossibility. One mission might require a probe to fit a specific engine bell. Another might need a robotic arm with a custom-made gripper. A third, for debris removal, might need a harpoon or a net. This technical problem proves the economic problem. There is no scalable, repeatable service, and therefore there is no profitable, viable business.
These technical challenges are not three separate problems. They are a “risk triad” – a set of interconnected, sequential, and multiplying failure points. A failure in any one phase guarantees the catastrophic failure of the entire mission. A normal space mission, like a launch, has one major, high-risk phase. An in-orbit servicing mission has at least three, and they happen in sequence:
- Phase 1: RPO Failure. The servicer fails to approach the target correctly. Like DART, it suffers a software bug, uses too much fuel, and collides, failing the mission before the main event even begins.
- Phase 2: Capture Failure. The RPO works, but the target is non-cooperative and tumbling. The robotic arm fails to get a secure grip. It might break off a fragile solar panel, damage the client, and send the servicer itself spinning away, out of control.
- Phase 3: Manipulation Failure. The capture magically works, but the servicing fails. The robot’s “contact dynamics” are wrong. A bolt that was never meant to be turned in a vacuum is frozen solid. The fuel line doesn’t form a perfect seal, and high-pressure propellant sprays into space.
The probability of mission success is not the average of these three risks; it’s the product of their success. If each of these three high-risk phases has a 10% chance of failure – a wildly optimistic assumption – the mission’s total chance of failure is not 10%, but nearly 27%.
The DART mission failed at Phase 1. A non-cooperative servicing mission, which is infinitely more complex, would be far more likely to fail at Phase 2 or 3, after making contact. This is the most dangerous possible outcome, as it guarantees the creation of a new, massive debris field.
Creating the Very Problem It Claims to Solve
The most significant and damning irony of the in-orbit servicing argument is its relationship with orbital debris. Servicing is promoted, first and foremost, as a “green” technology for space, a key solution to the “space junk” problem. Servicers are described as “garbage trucks” or “tow trucks” that will clean up the orbital highways.
A critical analysis shows that in-orbit servicing is far more likely to exacerbate this problem. It risks creating new, catastrophic debris fields, and its core business model encourages the cluttering of valuable orbits with “zombie” satellites.
The Risk of Catastrophic Debris Generation
Proponents claim servicers will act as “garbage trucks” to clean up LEO. This is a dangerously optimistic projection that ignores the immediate, overwhelming risk: a failed servicing mission is a debris-creating event.
A failed mission is not a quiet failure. It is a high-velocity collision between two large, intact, and often propellant-filled space objects.
The modern orbital debris environment, and the “Kessler Syndrome” scenario that haunts the space industry, is not the result of a million small failures. It is the result of two catastrophic events: the 2007 Chinese anti-satellite (ASAT) test and the 2009 collision between a dead Kosmos satellite and an active Iridium satellite. These two single events created thousands of new pieces of lethal, trackable debris, and millions of smaller, untrackable pieces. They single-handedly polluted entire orbital altitudes for decades.
A failed RPO – like the DART collision, but at a higher velocity and with larger objects – would be a near-perfect replication of these events. It would turn two multi-ton satellites into a “debris cloud” of shrapnel, shotgun-blasting through a valuable orbit and threatening every other satellite in that path.
The risk is so high that the NASA Office of Inspector General (OIG) has explicitly stated in its guidelines that servicing missions “should not generate debris.” This is a clear statement of the risk, but it is a hope, not a technical guarantee. Given the high probability of failure (as detailed in the “risk triad”), a complex robotic servicing mission is not a debris solution; it is one of the most significant debris-generation risks in orbit today.
Servicing Legacy Satellites: The “Zombie” Problem
This analysis challenges the very goal of servicing. The celebrated aim is to take a 15-year-old satellite that has run out of fuel and give it another 5 years of life, allowing its operator to extract another half-billion dollars in revenue.
This is not sustainability; it is the “zombie” problem.
You are spending hundreds of millions of dollars on a high-risk mission to “save” a satellite that is, for all intents and purposes, obsolete. A 15-year-old satellite has degraded solar panels, aging batteries that hold less charge, and computers with less processing power than a 10-year-old smartphone.
This is a bad investment for the orbital environment. It’s a short-term, profit-driven decision that keeps an underperforming, aging piece of junk in a highly valuable and limited orbital “slot.” This is not “sustainability”; it is “hoarding.” It actively prevents a new, next-generation, more-capable satellite from using that same precious orbital real estate.
The Servicer’s End-of-Life: The Ticking Time Bomb
The most significant and most overlooked debris risk associated with this new industry is the servicer itself. What happens to the “garbage truck” when it runs out of fuel or its own mission ends?
A servicing vehicle, by definition, is a new and uniquely dangerous class of space debris. To do its job, it must be large. It must be complex. And it must be full of propellant, high-pressure tanks, and complex electronics. When it dies, it will become one of the most dangerous potential “debris objects” ever launched.
A single impact from a small, existing piece of debris could cause a dead servicer, with its leftover propellant, to explode, creating a massive new debris cloud. It is a ticking time bomb.
What is the end-of-life plan for these servicers? The documentation for Northrop Grumman’s MEV states it has a 15-year service life. Its plan, at the end of a mission, is to simply undock and proceed to its next client. The documentation does not detail its final disposal plan.
The implied plan, and the one used by the MEV for its first client, is the “graveyard orbit.” This is an altitude a few hundred kilometers above the valuable Geostationary Orbit, where operators dump their dead satellites.
This “graveyard orbit” is not a solution. It is a short-sighted abdication of responsibility that creates a future, and potentially far worse, debris problem. It is sweeping the trash under a different rug.
First, these orbits are not perfectly stable. Over long timelines, gravitational perturbations from the sun and moon can push these dead objects back down, interfering with operational orbits once again.
Second, this practice concentrates the risk. We are now actively filling these graveyard orbits with a high density of large, dead, propellant-filled “zombie” satellites and their “time bomb” servicers. We are creating the perfect condition for a future Kessler Syndrome event, but in an orbit from which we can never recover.
This is the final, and greatest, irony. The “disposable” LEO constellation model, which is so often criticized, is, paradoxically, more environmentally sustainable. Those satellites are designed for a 100% controlled, atmospheric burn-up on re-entry. The “servicing” model, which claims sustainability, deliberately and permanently leaves massive, explosive pieces of junk in orbit.
The Unspoken Threat: Geopolitical Destabilization
Beyond the flawed economics, the obsolete business model, and the extreme technical risks, the most compelling case against in-orbit servicing is its geopolitical implication. A servicing satellite is not just a commercial tool. It is, by its very nature, a “dual-purpose” technology. Its development and deployment are not merely a commercial endeavor; they are a primary, if unspoken, driver of a new, subtle, and significantly dangerous space arms race.
Dual-Purpose, Not Dual-Use: A Weapon in Sheep’s Clothing
To understand the threat, it is essential to understand the critical distinction between “dual-use” and “dual-purpose.”
A “dual-use” system, like the Global Positioning System (GPS), has both civilian and military functions. It is designed to provide navigation for your car’s mapping app, and it is also designed to provide targeting data for a guided missile. Both uses are intentional.
A “dual-purpose” system is different, and far more insidious. It is a technology that is designed for a completely benign function – like repair, refueling, or debris removal – but which has characteristics that allow it to be repurposed for a hostile one.
An in-orbit servicing vehicle is the quintessential dual-purpose weapon.
A servicer equipped with dexterous robotic arms to “repair” a satellite can use those same arms to harm a satellite – to rip off a solar panel, crush a sensor, or break an antenna. A servicer designed to “refuel” a client can just as easily “poison” it. A servicer with a harpoon or laser to “remove debris” can use that same harpoon or laser on an active, operational satellite. A servicer that can “relocate” a dead satellite to a graveyard orbit can, by definition, “relocate” an adversary’s critical satellite into a useless orbit, effectively killing it without firing a shot.
A servicing vehicle is an anti-satellite (ASAT) weapon. It is a “weapon in sheep’s clothing.”
This technology provides a nation with a powerful, “plausibly deniable” first-strike capability. A country with a “civilian” space agency operating a fleet of “debris removal” or “servicing” vehicles could, at a moment of its choosing, employ that fleet as a weapon against an adversary’s space assets. The attack could be blamed on a “software glitch” or a “botched repair attempt.” This is not a “bug”; it is a core feature of the technology.
The Impossibility of Verifying Intent
This dual-purpose nature creates an intractable security dilemma that threatens to destabilize the entire orbital environment. In the “congested” and “contested” domain of space, it is impossible to distinguish a benign servicer from a hostile one until it’s too late.
This creates an unanswerable question for military operators, like the U.S. Space Force. What is the correct response when a Chinese or Russian servicing vehicle – from a nation with a known “military-civil fusion” doctrine – begins to approach a $5 billion U.S. national security satellite? What is the protocol when anycommercial servicer – which could be hacked or coerced by another government – approaches?
The International Committee of the Red Cross (ICRC) has warned of “harmful in-orbit rendezvous and proximity operations.” A “nonconsensual rendezvous,” even if it’s just an “inspection,” would be seen as an “indicator of intent to harm.”
This is the core of the security dilemma. By the time you have verified that the approaching servicer’s intent is hostile, it is already at point-blank range, and your multi-billion-dollar asset is lost. This ambiguity doesn’t raise the threshold for conflict; it lowers it. It may force a satellite’s operator to take defensive and escalatory action – like jamming, moving, or even firing a counter-weapon – against any object that comes close.
The development of in-orbit servicing technology is not simply risking a new space arms race. It is the new space arms race, cleverly disguised as a commercial and environmental enterprise.
A traditional ASAT, like a ground-launched missile, is a “kinetic” weapon. It is loud, obvious, and an undeniable act of war. It also creates a massive, self-damaging debris cloud, as the 2007 Chinese test proved. It is a blunt instrument.
A servicer is a far more sophisticated and destabilizing weapon. It is “plausibly deniable.” It is “non-kinetic” or “soft-kill.” It doesn’t need to create a debris cloud. It can simply grab an adversary’s satellite and gently push it out of orbit. It can “inspect” it to steal its secrets. It can “obstruct” its sensors, “hack” its systems, or “blind” its cameras. And it is already in orbit, a weapon-in-waiting, able to strike with zero warning.
This is the true, unspoken driver for this technology’s development by major space powers. It is not about “sustainability.” It is about “space superiority.” Its very existence forces all other nations to develop their own, escalating a new, costly, and incredibly dangerous arms race defined by ambiguity and suspicion.
A Framework of Failure: The Legal and Regulatory Void
The final, and perhaps most insurmountable, case against in-orbit servicing is that the entire concept is built on legal and regulatory quicksand. The international and national laws that govern space are relics of the Cold War, signed in the 1960s and 70s. They were designed for a world where only two superpowers launched simple “one and done” objects. They are completely and catastrophically unprepared for the complex, high-risk, multi-party commercial interactions of in-orbit servicing.
This legal vacuum is not an academic problem. It is the final link in the chain, cementing the economic, technical, and geopolitical failures, and making the entire business uninsurable and uninvestable.
The 1972 Liability Convention: Unfit for Purpose
The entire global framework for space liability rests on the 1972 Liability Convention. This treaty establishes two kinds of liability. If a rocket part falls from the sky and hits your house, the launching state is “absolutely liable.” But for damage that occurs in orbit – a satellite-on-satellite collision – the standard is different. The launching state is liable only if it was “at fault.”
This “at-fault” standard is an unworkable, un-litigable nightmare for in-orbit servicing.
As legal experts have pointed out, servicing missions, by their very nature, involve “two different space objects” and a “higher risk of damage.” If a collision occurs during a “capture” phase, it would be nearly “impossible to establish which one was at fault.”
- Was it a software bug in the servicer’s autonomous RPO system? (The servicer’s fault).
- Was it a design flaw in the client’s non-cooperative engine bell, which broke off when grabbed? (The client’s fault).
- Was it an un-tracked piece of millimeter-sized debris that hit the servicer’s camera at the last second, causing it to miscalculate? (No one’s fault).
This legal “black hole” for liability creates a level of paralyzing uncertainty. It poses a massive, unquantifiable risk for the launching states (who, under the Outer Space Treaty, are ultimately responsible for the actions of their private companies). And it is a source of absolute dread for commercial investors.
Non-Consensual Proximity and Sovereignty
The law is not just old; it is silent on the very act that defines servicing: approaching another nation’s satellite without its permission.
There are no internationally recognized “rules of the road” for space. There is no law that specifies how close a “non-consensual” approach can be. There is no “keep-out” zone.
This legal ambiguity is a geopolitical flashpoint. Under the Outer Space Treaty, states are responsible for the activities of their private companies. If a “rogue” U.S. commercial servicer, in an attempt to “inspect” a failed satellite, approaches a Chinese military satellite, China will not see it as a commercial act. It will see it as an act of aggression by the United States government. This ambiguity blurs the line between a commercial service and an act of war.
Paralyzing the Market (Revisited)
This is where the entire “case against” comes full circle. The economic, technical, and geopolitical flaws are all cemented in place by this legal and regulatory void, which paralyzes the market.
Investors are, as one legal analysis noted, “skeptical of high-risk endeavors that lack consistent government oversight.”
The lack of regulation means an investor, a bank, or an insurance underwriter has no answer to the most basic, fundamental question: “Who pays when this goes wrong?”
The market for in-orbit servicing hasn’t failed because the technology is hard, though it is. It has failed because the legal, geopolitical, and insurance risks are so significant, so un-priceable, and so interconnected that no rational investor will fund it, and no government can safely permit it.
Summary
The vision of in-orbit servicing – a sustainable, circular ecosystem of repair and refueling in space – is intuitively powerful. It draws on familiar, logical analogies of repair and maintenance. But a critical, objective examination of the concept reveals it to be an orbital mirage, an idea that is not just unworkable but also undesirable.
The business case for servicing is fundamentally broken. It is a high-cost, high-risk solution for which there are few, if any, willing customers. The economics are built on a 50% cost-of-replacement rule that servicing missions cannot meet, and the entire venture is permanently hamstrung by a “chicken and egg” market failure that incumbent manufacturers have no financial incentive to solve. Investors, wary of the unproven technology and the lack of a clear customer base, have rightly remained skeptical.
Even if the economics were sound, the servicing model is already being made obsolete by a superior and truly revolutionary business model: the “disposable” LEO constellation. When satellites are mass-produced for under two million dollars, the idea of launching a multi-hundred-million-dollar robotic mission to repair one is economically absurd. This new paradigm – based on rapid replenishment, not risky repair – delivers superior resilience and technology upgrades far more cheaply and reliably than servicing ever could.
Operationally, servicing remains a nightmare. The autonomous rendezvous and docking required for these missions are exceptionally difficult and prone to catastrophic, debris-creating failure, as demonstrated by past mishaps. The vast majority of potential targets are “non-cooperative,” meaning they are tumbling, silent, and not designed to be grappled, turning every mission into a high-stakes, custom R&D project.
Ironically, the servicing and debris-removal industry risks becoming a primary source of new orbital debris. A failed mission could create a dense debris cloud, and the servicers themselves, at their end of life, will become some of the largest, most dangerous, and most toxic “time bombs” ever left in orbit.
Finally, the entire endeavor is a source of significant geopolitical instability. A servicer’s “dual-purpose” capabilities make it indistinguishable from an anti-satellite weapon. The technology, far from promoting sustainability, is a key driver in a new, destabilizing space arms race, providing a cloak of “plausible deniability” for hostile acts. This, in turn, has created a legal and regulatory void. The 1970s-era liability laws are utterly unfit for purpose, making the business uninsurable and the market uninvestable.
The case for in-orbit servicing has failed. It is an elegant solution to a problem that no longer exists, and in its pursuit, it creates new and far more dangerous economic, environmental, and military threats.
