AI-generated conceptual illustration of a generic spacecraft near a binary asteroid, with the amber enclosure representing a protected guest-software sandbox. It does not depict Hera’s actual design or a selected experiment.
On August 5, 2026, the European Space Agency invited European researchers and companies to propose software experiments that could run aboard Hera after the spacecraft completes its main asteroid investigation.
ESA’s announcement describes a carefully limited second use for the planetary-defence spacecraft: a software laboratory about 150 million kilometres from Earth. The agency is considering experiments in onboard intelligence, image processing, and artificial-intelligence-based operations and autonomy, with a month of operations planned for August 2027.
The proposal remains contingent on Hera completing its primary work, ESA selecting suitable ideas, and chosen teams submitting their implementation and source code for validation. No outside software has yet been announced as selected, approved or ready to fly.
A mission built around an asteroid-deflection test
Hera launched in 2024 and is expected to reach the Didymos–Dimorphos asteroid system in autumn 2026. Its primary task is to examine the aftermath of NASA’s Double Asteroid Redirection Test, or DART, and improve scientists’ understanding of whether a spacecraft impact could help deflect a hazardous asteroid.
DART struck Dimorphos on September 26, 2022. NASA later confirmed that the impact changed the small asteroid’s orbit around Didymos. Neither asteroid posed a threat to Earth; the collision was a controlled planetary-defence demonstration.
ESA expects Hera’s main investigation around the two asteroids to finish by mid-2027. The proposed software campaign would come afterwards, making use of a spacecraft already operating in deep space rather than changing the objectives of its primary mission.
What the software lab would—and would not—do
Hera’s onboard computer uses a European-developed dual-core LEON3 processor. Under ESA’s plan, one processor core would continue operating the spacecraft. Guest applications would run separately on the second core inside a protected “sandbox,” an isolated environment intended to keep experimental code apart from the spacecraft’s normal operations.
ESA says access to instruments and subsystems would be facilitated for the experiments. Individual software runs would be limited to about two to three hours at a time and stopped immediately if any problems were identified.
The agency has been explicit that the goal is not to hand control of Hera to outside software. The proposed arrangement would instead create limited test periods in which validated guest code could run alongside the mission’s established systems.
“Protected” should not be read as risk-free. It describes ESA’s planned isolation and shutdown measures; the announcement does not report completed guest-software tests or prove that every proposal can be accommodated safely.
The ideas ESA is looking for
ESA lists onboard intelligence, image processing and AI-based operations or autonomy as possible experiment areas. The broad aim is to explore whether more processing and decision support aboard future spacecraft could reduce reliance on frequent ground oversight and make missions more resilient.
Those are programme goals, not capabilities that Hera has already demonstrated. “AI” in this announcement does not mean an AI system would independently command the spacecraft, nor does the call establish that any proposed approach will work in flight.
The importance of the plan lies in the setting. ESA is proposing to test selected software aboard an operational spacecraft in deep space, after its main asteroid work, rather than announcing another ground demonstration. What can actually be learned will depend on which ideas survive review, what ESA validates, and how the limited runs perform.
How the proposal path is scheduled
Applicants first submit an outline through ESA’s Open Space Innovation Platform, or OSIP, for an initial evaluation. The invitation is directed at European researchers and companies; applicant-specific eligibility and participation conditions still have to be checked through ESA.
ESA says it plans to select ideas in mid-October 2026. Chosen teams would then be invited to provide their implementation package and full source code for detailed validation ahead of the planned month of operations in August 2027.
That sequence matters. Initial selection would not by itself amount to approval to run software aboard Hera. Validation remains a later step, and the software campaign also depends on the spacecraft completing its primary planetary-defence work.
What to watch next
The next meaningful milestone is ESA’s first-stage selection in October 2026, followed by validation of the chosen implementation packages. If the schedule holds, Hera could move from investigating humanity’s first asteroid-deflection test to hosting a tightly controlled software experiment for future deep-space missions.
For now, the verified development is narrower: ESA has opened a proposal process and described the limits it intends to place around any guest software. Claims about successful autonomy, approved experiments or flight results would be premature.
