Link to the code: Zae-Project / arkspace-core

On-Orbit Servicing in 2026: ELSA-M, Tetra-5, and the ASSET Toolkit

Four separate on-orbit servicing missions are scheduled within a twelve-month window in 2026. That density of activity has not happened before. The question the industry has been asking for a decade, whether satellite servicing can become a reliable commercial service rather than a one-off demonstration, is getting answered this year with mission hardware rather than feasibility studies.

The Economic Problem Servicing Solves

About 80 percent of geostationary satellites that reach end-of-life still function electronically. They run out of propellant, not capability. Every year, 10 to 20 GEO satellites are retired that could otherwise provide years of additional service if someone could deliver fuel or push them back to their operational slot.

In low Earth orbit the economics are different but the underlying logic is similar. LEO satellites are cheaper and shorter-lived, but the growing constellation scale creates a different problem: defunct satellites accumulate faster than they deorbit naturally. A dead satellite in a high LEO shell (above 600 km) can take decades to re-enter on its own. Removing it actively requires a dedicated mission. As of 2026, the scale of uncontrolled debris in LEO has made active removal no longer optional for sustainable constellation operation.

On-orbit servicing addresses both: extending the productive life of functional but fuel-depleted assets, and removing debris that represents collision risk.

ELSA-M: Magnetic Capture for Commercial Debris Removal

Astroscale’s ELSA-M (End-of-Life Service by Astroscale, Multi-client) is the highest-profile LEO servicing mission of 2026. The 600 kg spacecraft uses a magnetic docking system to capture satellites equipped with a compatible docking plate. The target: a OneWeb telecommunications satellite in a 1,200 km orbit.

The mission architecture requires preparation on the client satellite side. A magnetic docking plate, small and light enough to add to a standard satellite bus during manufacturing, is the interface that enables ELSA-M’s capture mechanism. This is not a grappling arm or net approach. It is a standardized interface system, analogous to what a standardized docking port does for crewed spacecraft.

ELSA-M’s launch was confirmed in March 2026 through a launch service agreement with Isar Aerospace, using Isar’s Spectrum rocket from Andøya Space in Norway. The in-orbit demonstration is scheduled to capture and de-orbit the OneWeb satellite in 2027. Funding is primarily private (Astroscale) with co-funding from the UK Space Agency through ESA’s Connectivity and Secure Communications Directorate.

This follows Astroscale’s 2021 ELSA-d demonstration, which validated the magnetic capture and autonomous rendezvous system. ELSA-d simulated debris capture. ELSA-M does the real thing on a real defunct satellite. TRL has moved from 6 (system demonstration in relevant orbit) toward 7–8 (operational mission).

The autonomous rendezvous and docking capability involved is directly related to the spacecraft AI capabilities covered in Autonomy Ascending: How Spacecraft Are Learning to Fly Themselves in 2026. Autonomous approach and close-proximity operations are harder than attitude control or orbital maneuvering — they require real-time perception and decision-making with no communication round-trip to ground.

Tetra-5 and Tetra-6: Standardizing Refueling

While Astroscale works the debris removal problem in LEO, the refueling ecosystem is being standardized in GEO and cislunar space through the US Space Force’s Tetra program.

Tetra-5 launches in June 2026. The mission deploys two small satellites equipped with Orbit Fab’s Rapidly Attachable Fluid Transfer Interface (RAFTI), a standard fueling port designed to be installed on any satellite during manufacturing. One satellite docks with Orbit Fab’s propellant depot, the other tests compatibility with Astroscale’s propellant shuttle. In the full mission sequence, the Astroscale spacecraft refuels a Space Force satellite, refuels itself from the Orbit Fab depot, then refuels a second Space Force satellite. Orbit Fab’s depot launches on the same vehicle.

Tetra-6 follows in 2027, demonstrating Northrop Grumman’s competing Passive Refueling Module (PRM) docking interface using the ROOSTER-5 (Rapid On-orbit Space Technology Evaluation Ring) tanker.

The competition between RAFTI and PRM matters because refueling standardization requires the industry to converge on one or a small number of interface standards. A satellite built today cannot be refueled unless it was designed with a compatible interface. The Tetra missions are funding the standardization race, much the way USB standards emerged from competing solutions.

Northrop Grumman already demonstrated life extension with its MEV (Mission Extension Vehicle) in GEO. MEV-1 docked with Intelsat IS-901 in February 2020, the world’s first commercial satellite life extension mission. MEV physically docks to client satellites and provides propulsion for repositioning and station-keeping. The Mission Extension Pod, a smaller derivative, is on track for launch in 2026.

The ASSET Toolkit: Engineering the Servicing Spacecraft

The AIAA SciTech Forum 2026 introduced the Automated Satellite Servicing Engineering Toolkit (ASSET), a modular framework for designing servicing spacecraft in the 200–300 kg class.

ASSET addresses a specific engineering gap: there is no standardized design reference for building a servicing vehicle. Every prior mission (MEV, ELSA-d, OSAM-1) was bespoke, which drives development cost and timeline. ASSET provides a modular architecture where subsystems for rendezvous, docking, propellant handling, and inspection can be configured for specific client satellite geometries and mission requirements.

A 200–300 kg servicing spacecraft is a practical size range for both LEO and GEO missions. Light enough to be a secondary payload on many launchers, capable enough to carry meaningful propellant mass or robotic arm systems. ASSET essentially defines what a standard space mechanic looks like.

The TRL for ASSET as a toolkit is 2–3 (concept formulated, analytical and experimental validation in progress). Its value is shortening the path from mission concept to flight hardware for the next generation of operators.

Why This Matters for Orbital Infrastructure at Scale

The constellation-scale orbital infrastructure proposals described elsewhere, SpaceX’s 1 million satellite data center filing, China’s Three-Body constellation, ESA’s ASCEND, all assume satellites that operate for years and can be maintained or replaced efficiently.

Active debris removal capability changes the risk calculus for large constellations. If defunct satellites accumulate and create collision cascades, the operating environment for any constellation degrades. Serviceable constellations, where satellites are designed with standard refueling ports and docking plates from the outset, have a fundamentally different end-of-life profile than today’s “deorbit or drift” model.

The radiation-hardened processor work described in Carnegie Mellon’s radiation-hardened chip development is relevant here: a servicing spacecraft that operates autonomously at high radiation levels for extended missions needs exactly the kind of compute the CubeSat program is testing. On-orbit servicing vehicles are among the most demanding applications for radiation-tolerant autonomy.

The edge AI processing advances in operational smallsats also feed into the servicing problem. A servicing spacecraft that can visually identify a target satellite’s docking interface, plan a rendezvous trajectory, and execute close-proximity maneuvers autonomously is a direct application of the edge inference pipelines that D-Orbit and others are developing for Earth observation.

Path Forward

The four 2026 missions (ELSA-M, Tetra-5, the MEV pod launch, and the Robotic Servicing of Geosynchronous Satellites demonstration) represent the largest single-year concentration of on-orbit servicing activity in history. If they proceed without major failures, the outcome is a validated set of commercial business models for debris removal, life extension, and refueling that operators can use when procuring future satellites.

The longer-term goal is that on-orbit servicing becomes a standard line item in satellite constellation planning, not an exceptional response to stranded assets. A constellation designed from the outset around a servicing architecture could operate at higher density with lower debris risk than today’s designs. That changes the economic model for orbital infrastructure significantly.

The docking plate standardization effort (RAFTI vs. PRM) will take years to resolve. But the fact that the US Space Force is funding competing standards rather than waiting for industry consensus suggests that policymakers have concluded that standardization is worth accelerating through procurement pressure.

Official Sources