Where Does Data Wait When Contact Ends? CAPSTONE's Lunar Networking Test

A small spacecraft near the Moon is sending information to Earth. Before the transfer finishes, the communication opportunity ends. The immediate engineering question is not simply how powerful the antenna should be. What happens to the information that has not yet made it across? In a CAPSTONE demonstration described by NASA in July 2026, the spacecraft kept the remaining data and automatically resumed transmission at the next opportunity. All the data in that demonstration reached Earth. The achievement was not preventing an interruption, but preventing the interruption from permanently breaking the transfer. [1]

That event illustrates a choice in the design of space communications. Instead of treating an available connection as a continuous background condition, a system can incorporate periods without contact into its ordinary operation. CAPSTONE tested this approach near the Moon after its original objectives had led to an extended role as a software experimentation platform. This account uses NASA reports published before the knowledge cutoff of 13 October 2026. A successful demonstration is evidence of a particular capability, not an unlimited guarantee for every future mission. [1]

A second job for hardware already in flight

Launched in June 2022, CAPSTONE was approximately the size of a microwave oven. Terran Orbital built the spacecraft, which was owned and operated by Advanced Space. A central early objective was to characterize operations in a near-rectilinear halo orbit, where the combined gravitational influence of Earth and the Moon shapes the path. Flying and characterizing that orbit supplied operational experience beyond a calculated trajectory. It did not, by itself, answer every question about communications or navigation. [1][2]

During its extension, the same spacecraft became a host for additional communications and autonomous-navigation experiments. NASA's 2026 account emphasized that existing hardware could accommodate new applications after launch, without sending another satellite. The significance was not merely that software could be installed. The applications were running within the contact opportunities and operational constraints of an actual lunar environment, rather than only on a computer on the ground. [1]

The resulting mission was therefore not one function becoming progressively more powerful. It was a platform on which distinct questions could be tested. How would a spacecraft estimate its position? How would information reach its destination? What could continue when Earth contact became scarce? A single vehicle can host all these investigations while each still requires its own evidence of success.

Storage is not a radio connection

Delay/disruption tolerant networking, or DTN, treats waiting as part of information transport. Data can remain aboard the spacecraft while a connection is unavailable and be forwarded when contact returns. In the reported CAPSTONE demonstration, a transmission began, the connection ended before completion, and the remaining data stayed aboard until transmission automatically resumed at the next opportunity. The observed result was continuity of delivery across a discontinuity in contact. [1]

There are two different kinds of waiting to distinguish. A signal takes time to travel through space. Information may also have to wait before a suitable communication opportunity becomes available. Storing and forwarding helps manage the latter situation; it does not remove signal propagation time. Nor does holding information in memory make a radio path available at that moment. Resilience to interrupted contact is different from uninterrupted connectivity.

Imagining a surface vehicle temporarily hidden by terrain can help explain why the principle matters. It must not be mistaken for the experiment that CAPSTONE actually conducted. NASA described a spacecraft-to-Earth transfer interrupted before completion and resumed later, not a rover driving behind a particular lunar hill. Applying the approach to another vehicle still involves that vehicle's storage, power, radio system, and operational conditions. [1]

The short mission report does not provide a numerical storage capacity, a maximum retention period, or a complete account of every possible failure. It would be unjustified to turn the successful transfer into a guarantee of delivery after any interruption. Keeping data does not create unlimited memory or energy. Hardware loss and exhausted power are not failures shown to have been solved by this demonstration. Those distinctions follow from the scope of the reported test, rather than from additional performance measurements.

Knowing a position is a different task

Another CAPSTONE experiment concerned autonomous Navigation, Guidance, and Control software, abbreviated autoNGC. Its purpose was to help a spacecraft determine where it was, where it was going, and how to proceed without waiting for instructions from the ground. Portions of the software had previously flown in Earth orbit; CAPSTONE provided its first test at the Moon. Periods with only a few Earth-contact passes per week became an opportunity to evaluate onboard operation. [1]

The spacecraft's star tracker camera supplied images of the Moon, Earth, and other celestial bodies for optical navigation. NASA reported that, at times, the camera-based approach outperformed ground-based methods for real-time onboard navigation. The qualification matters. The report did not establish superior accuracy in every situation, provide a universal position-error figure, or show that all ground support had become unnecessary. [1]

DTN and optical navigation can consequently be useful together without doing the same job. One manages interruptions while transporting information. The other uses observations to help estimate the spacecraft's state. Successful delivery of a data set does not demonstrate that every maneuver during the intervening period was performed independently and without error. Keeping these functions separate makes the accomplishment more intelligible, not less impressive.

An earlier radio experiment had another meaning

CAPSTONE's May 2023 navigation test provides a further useful distinction. On 9 May, it sent a signal to NASA's Lunar Reconnaissance Orbiter. LRO returned the signal, which CAPSTONE converted into a measurement. The exchange was designed to obtain distance and relative-velocity information for the Cislunar Autonomous Positioning System, or CAPS. NASA's contemporary report described the demonstrated ability to collect measurements that the positioning software could use. [2]

That exchange was not the same measurement method as camera-based optical navigation. Neither was it the same function as retaining data during an interrupted transfer. Combining all three under the word autonomy can hide the engineering chain: producing an observation, estimating a state from observations, and transporting information to another endpoint are distinct activities. Each can be investigated in flight, with a different definition of what counts as a successful test.

NASA's activities aboard CAPSTONE concluded in June 2026. Its July report stated that Advanced Space would continue using the spacecraft as a technology development testbed. That statement should not be rewritten as a spacecraft shutdown. What remains especially informative is the operational evidence: an interrupted transfer resumed, stored information was delivered, and onboard techniques were evaluated under limited contact. A lunar network is judged not only by what it does while speaking, but also by how it carries information through the intervals when it cannot. [1]

Sources

[1] Powers, K. NASA (6 July 2026). NASA’s CAPSTONE Completes Extended Mission Testing Lunar Technologies. https://www.nasa.gov/technology/space-comms/nasas-capstone-completes-extended-mission-testing-lunar-technologies/

[2] Tabor, A. NASA (18 May 2023). CAPSTONE Takes Moon Shot, Successfully Tests Navigation Technology. https://www.nasa.gov/image-article/capstone-takes-moon-shot-successfully-tests-navigation-technology/