Satellite overcrowding around Earth: How AI can prevent collisions, interference with intelligent 'space autopilot'
Space has never been so crowded. As of March 2025, there were more than 14,000 individual satellites in orbit, in addition to 27,000 tracked objects counting inactive satellites, rocket stages and large debris, marking a 31% increase over 2023, thus confirming the exponential growth of space traffic. This is alarming to the scientific community. As overcrowding increases, so does light pollution (which makes astronomical observation more difficult) as well as the risk of collisions, cyber-attacks and interference in communications.
In response to the challenges of satellite overcrowding, universities and companies are collaborating to find innovative solutions based on artificial intelligence. This is how ASIMOV was conceived, an intelligent "space autopilot" capable of autonomously approaching, mapping and monitoring inactive or non-cooperative objects (such as failed satellites) for inspection, maintenance or removal.
The project is coordinated by AIKO, a Turin-based scaleup that develops advanced software centered on artificial intelligence and automation for space applications, in partnership with Politecnico di Milano, T4i and Tiny Bull Studio.
"The problem of space overcrowding is complex. Hence, tackling it requires a common approach, which is why collaboration between space companies, universities and research institutes is essential. It is only by converging different competences that we can tackle this challenge concretely to offer effective answers at an international level, developing innovative solutions such as the space autopilot.
"Our ultimate goal is to build a sustainable orbital future in which we protect the environment and preserve access to space for generations to come," says Lorenzo Feruglio, CEO and Co-founder of AIKO.
"We want to show that artificial intelligence is a key enabler for optimizing satellite monitoring and management. It allows us to map and control operations that were unthinkable until recently. AI is not just an automation tool. It is a technology capable of increasing the reliability and safety of missions, while minimizing risks and inefficiencies."
"The Politecnico di Milano contributes to the development of algorithms, and plays a central role in hardware-in-the-loop verification activities of the navigation and autonomous driving techniques developed by project ASIMOV," says Michèle Lavagna, Full Professor of Flight Mechanics at the Politecnico di Milano.
"In the ARGOS laboratory, we reproduce orbital dynamics by coordinating robotic arms, optical sensors and satellite models, capturing real-time images in an environment that mimics deep space. This new frontier of autonomous control of satellite systems is crucial for future proximity and in-orbit servicing operations."
The challenge of space overcrowding—The increasingly crowded low Earth orbit is posing an ongoing issue of rising risks. In addition to the danger of collisions between satellites, which can cascade in thousands of new debris, the phenomenon also has direct repercussions on scientific research and safety.
Indeed, "mega-satellite constellations" reflect sunlight, causing light pollution that disrupts astronomical observations and impairs the discovery of crucial phenomena, such as the passage of asteroids. This is compounded by interference in communications, risks to astronauts, and vulnerability to possible cyber or physical attacks. Inactive satellites and abandoned rocket stages also contribute to the problem. Without disposal or active removal solutions, they end up wandering uncontrolled in space, turning from technological resource to orbital threat.
Solutions: the average lifetime of a satellite and how to dispose of it—The lifetime of a satellite varies according to its mission, type and orbit. The smallest ones will have a lifetime of a few years, while large or more distant ones can remain operational for decades. Function is also a decisive factor.
While some satellites circling in low Earth orbit (LEO) naturally re-enter the atmosphere and destroy themselves, others remain wandering in space, generating potentially dangerous debris. Hence, disposal strategies and "cleaning missions" are being developed today.
There are satellites assigned to remove objects that are no longer active, while others are equipped with refueling and life extension systems, as well as controlled de-orbiting technologies for devices that have reached the end of their mission. These are complemented by international guidelines and programs, such as ESA's Zero Debris Charter, promoting more sustainable design and shared responsibility in space traffic management.
ASIMOV, the space autopilot for the inspection of non-cooperative objects—designed to address the growing problem of debris and inactive satellites, ASIMOV (Autonomous System for In-orbit Mapping and Observation of non-cooperative Vehicles) is a project coordinated by AIKO, in collaboration with Politecnico di Milano, Technical University of Munich, T4i and Tiny Bull Studio. The aim is to build the first truly intelligent "space autopilot" capable of autonomously inspecting and mapping non-cooperative objects in low Earth orbit, such as failed satellites or unknown debris.
The project's core unit is an innovative Guidance, Navigation and Control (GNC) system based on artificial intelligence. It integrates autonomous navigation forms and Reinforcement Learning algorithms to help plan approach trajectories. These technologies will enable the satellite to recognize and reconstruct the geometry of objects to be inspected without support from Earth, thus enhancing the safety and sustainability of orbital operations.
A distinctive aspect of ASIMOV is its testing infrastructure. The ARGOS robotic facility at the Politecnico di Milano was upgraded to simulate orbital conditions in the laboratory, and to validate the systems developed. This dual approach—software development and physical testing—will lead to the creation of an operational prototype, besides being a strategic legacy for the future of the entire European space ecosystem.
Provided by Polytechnic University of Milan
