Starlink was born with an easy-to-understand promise: to offer fast internet from space where fixed and mobile networks don't reach or fall short. To achieve this, the company, founded by Elon Musk, has deployed thousands of small satellites in low Earth orbit, much closer to Earth than traditional geostationary satellites, which reduces latency and improves connection speed.
Over time, that initial approach has proven insufficient. SpaceX's constellation no longer aims to be simply a large "cable" descending from the sky. The current objective is to transform Starlink into an orbiting digital infrastructure , capable not only of transporting data, but also of managing and, in part, processing it directly in space, almost like a computer distributed around the planet.
From floating repeaters to a decision-making network
For decades, the typical image of a communications satellite has been that of a passive repeater : it receives a signal from Earth, amplifies it, and forwards it to another area. All the network intelligence—routing, control, traffic prioritization—has always been concentrated in ground stations, data centers, and network equipment located at ground level.
Starlink's approach breaks with that logic. Its constellation isn't designed as a collection of isolated satellites, but as a network of moving nodes . Each satellite travels at high speed, constantly changing neighbors, and yet must maintain stable connections. To achieve this, it relies on laser links between satellites that allow data to jump from one to another before reaching Earth.
This detail completely changes the role of space in internet architecture. Instead of simply sending information to the first available antenna, packets can travel along alternative routes within the constellation until they find the most convenient outlet. From a network perspective, the sky begins to function as a kind of global backbone that crosses oceans, remote areas, and regions with limited terrestrial infrastructure.
When a company controls a backbone network with these characteristics, the logical step is to stop selling only "basic connectivity" and start offering higher-level services . That's where this new phase of Starlink comes in, aimed at ensuring that the satellites don't just repeat signals, but actively participate in how data moves.
A very particular kind of edge computing in orbit
The idea of Starlink functioning as a "giant computer in space" might sound like science fiction, but in practice the concept is much more realistic. It's not about building massive data centers in orbit or training massive artificial intelligence models above our heads—something unrealistic due to energy consumption, heat dissipation, and technical complexity.
The company's goal is to move some of the digital tasks currently performed on the surface to the constellation. This approach is very similar to edge computing : moving intelligence to the edges of the network, where data is generated or where faster responses are needed, instead of centralizing everything in large processing centers.
In the case of Starlink, this “edge intelligence” would translate into capabilities such as prioritizing specific traffic flows , filtering redundant information, detecting anomalous behavior, and making routing decisions directly from the satellite. In this way, the constellation would cease to be a mere neutral conduit and become a platform that provides added value.
From a network perspective, this helps alleviate congestion, reduce response times, and prevent all data from passing through the same terrestrial infrastructure. For certain services, it also allows for the implementation of security or quality of service policies without relying exclusively on terrestrial nodes, which is especially valuable when operating on a global scale.
Beyond the “megabit”: services for critical sectors
In today's satellite market, simple connectivity is becoming a difficult product to differentiate . As more low-Earth orbit constellations appear with broad coverage and similar speeds, users are no longer just looking at the megabits per second figure, but at what they can actually do with that connection.
This is where sectors like commercial aviation, maritime transport, international logistics, and emergency services come into play , where the priority isn't simply having "internet access," but rather service continuity, advanced traffic management, and predictable latency even in challenging situations. These activities require communications that can withstand extreme environments and won't fail when they are most needed.
A constellation that makes decisions from space can adapt better to these scenarios. If a network segment becomes saturated in a region, the orbital infrastructure itself can reorganize routes, reserve capacity for critical services , or prioritize certain customers without waiting for constant instructions from Earth.
For the average home user, these changes will be less noticeable on a daily basis. What they will likely notice, if the model takes hold, is greater stability during peak hours, a more consistent response during demand spikes, and, in general, performance closer to that of a well-proportioned fiber or mobile operator , even if the signal comes from the sky.
The major limitation: energy, heat, and the laws of physics
Transforming a satellite into something more than just a repeater isn't simply a matter of updating the software . It involves redesigning the hardware itself and accepting a series of very clear physical limitations. In a terrestrial data center, if more computing capacity is needed, more servers are installed, cooling is reinforced, and more electricity is contracted. In orbit, that margin for maneuver simply doesn't exist.
A satellite operates on a fixed energy budget. All the energy it uses comes from solar panels and batteries , managed by systems that already have to power communications, position control, propulsion, and other vital subsystems. There's no power outlet or backup generator: what needs to be distributed is the energy coming in through the panels.
Added to this is the problem of heat. Every watt consumed is converted into thermal energy which, unlike on Earth, cannot be dissipated with fans or liquid cooling. In space, the only way is through thermal radiation in a vacuum , which necessitates the design of very carefully planned radiators, emitting surfaces, and conduction pathways.
The more computing power is added, the more demanding the thermal management system becomes, and the more complicated it is to keep the satellite within its operating limits. Ultimately, each unit's ability to "think" is directly tied to its energy budget and thermal design, leaving no room for improvisation once in orbit.
Managing energy as if it were software
In this context, the key lies not only in capturing more energy, but in treating it as a programmable resource . In many traditional satellites, power management is conceived as something relatively static: first ensure survival and control, and then use what's left over to power the payload.
In a constellation that aims to perform digital functions in orbit, energy becomes a dynamic budget . Certain processing tasks can be carried out when the satellite is well-lit and the panels generate more power, while in the shadowed sections of the orbit, consumption will have to be reduced and only the essentials for maintaining the mission prioritized.
This approach forces us to ask what is more “expensive”: spending watts transmitting data or processing it locally? In some cases, it can be efficient to dedicate energy to reducing the volume of information traveling over the network, for example by filtering duplicates, compressing, or summarizing data before downloading it to a ground station.
Therefore, the goal is not to send an excessive amount of computing power into space, but rather the right amount of computing power, well-suited to the type of services to be offered. The key is that this intelligence tangibly improves network performance without compromising the longevity or stability of the satellites.
From access network to global digital platform
If Starlink manages to pull off this shift, the change will not only be technological, but also in its business model. The constellation would cease to be seen as a simple system for providing internet access and would instead function as a platform for digital services in orbit , capable of offering advanced features directly from space.
This approach moves it away from the classic profile of a satellite operator and closer to a distributed digital infrastructure , equivalent to a large network of computers surrounding the planet. A key detail is that hardware in low Earth orbit is renewed relatively quickly because satellites have shorter lifespans and are continuously replaced.
This constant renewal allows for the incremental introduction of hardware improvements and new capabilities , almost as if we were updating a software-defined network. Each new batch of satellites can bring more efficient processors, better communication systems, or more advanced energy management mechanisms.
All of this points to a scenario in which the sky near Earth begins to function as a new layer of the global digital infrastructure , situated above terrestrial fiber and mobile networks, but intimately connected to them.
Europe and the challenge of coexisting with megaconstellations
While Starlink is moving in this direction, Europe is closely monitoring the development of low-Earth orbit megaconstellations . The European Union is working on its own initiatives to ensure its autonomy in satellite connectivity, but at the same time, it must regulate the presence of thousands of private satellites already operating over its territory.
For countries like Spain, with rural areas and difficult terrain where deploying fiber optics remains complex, solutions like Starlink offer a real alternative to closing the digital divide. However, the potential conversion of the constellation into a digital platform raises additional questions about technological dependence, interoperability with European networks, and the management of sensitive data.
Brussels and national authorities are also closely monitoring the impact on space traffic and orbital safety . As the number of satellites in low Earth orbit grows, so does the risk of collisions and the generation of fragments that could affect other missions, including European scientific and governmental ones.
The coexistence of commercial projects like Starlink and future European systems will require coordination agreements, common standards, and, presumably, stricter rules on how these satellites are launched, operated, and retired at the end of their useful life.
Space debris and the sustainability of the model
One of the side effects of this leap towards such dense constellations is the increased congestion in low Earth orbit . Each new satellite adds complexity to tracking space traffic and, in the event of a failure, can become debris that remains orbiting the planet for years.
The risk is not merely theoretical: a collision between satellites could generate a cloud of debris capable of endangering other missions , including manned vehicles, scientific platforms, or even other communication systems. Therefore, the discussion surrounding these new digital platforms also includes aspects of safety and long-term sustainability.
In parallel with technological evolution, international organizations, space agencies and operators are increasingly focused on defining best practices for satellite retirement , collision avoidance maneuvers and design requirements that reduce the risk of creating space debris.
The challenge for projects like Starlink will be to demonstrate that it is possible to maintain a complex and very large infrastructure in orbit without compromising the viability of the space environment for future generations.
What's at stake with this new phase of Starlink goes far beyond offering connectivity where fiber doesn't reach: the constellation aims to become an additional layer of the planet's digital infrastructure , with satellites that no longer just send and receive data, but understand it, organize it, and make decisions about it in flight, all under the strict rules of limited energy, physics, and an increasingly crowded orbit.
