Most of the people who converged on the Kennedy Space Center in mid-November for the third Artemis 1 launch attempt were focused on the giant Space Launch System rocket standing at Launch Complex 39B and the Orion spacecraft on top. Craig Hardgrove was thinking about something much smaller.
Hardgrove, a professor at Arizona State University, is principal investigator for LunaH-Map, one of 10 cubesats that hitched a ride on SLS as secondary payloads. The six-unit (6U) cubesat carried a neutron spectrometer designed to map water ice concentrations at the moon’s south pole.
In the months leading up to the launch, he had been quietly raising concerns about the health of LunaH-Map. The spacecraft was delivered in mid-2021 and installed on the rocket that fall, after which there was no ability to recharge its batteries. As the Artemis 1 launch slipped from early 2022 to late in the year, he worried that the batteries were discharging, keeping the spacecraft from operating immediately after deployment.
At the KSC press site hours before liftoff, he was cautiously optimistic about LunaH-Map. Ground tests of batteries like those on the cubesat showed a low discharge rate, suggesting they should still have plenty of charge left. Even if the batteries were depleted, he said the spacecraft’s solar panels could charge them up enough to get the spacecraft ready for a key maneuver days after launch.
He was right not to be concerned about the cubesat’s batteries. “Our batteries were at 70% state of charge when we got our first piece of telemetry,” he said in a talk about LunaH-Map at the Fall Meeting of the American Geophysical Union (AGU) a month after launch. “It was in line with our very optimistic predictions about where our batteries might be.”
The problem instead was with the cubesat’s propulsion system, an electric thruster called BIT-3 from Busek that uses solid iodine as propellant. The thruster did not operate as expected in the days after launch, causing the spacecraft to miss its primary opportunity to maneuver into orbit around the moon.
Hardgrove said telemetry from LunaH-Map suggested that a valve in the thruster was partially stuck closed. “The sticking is something that we knew about,” he said, suggesting it came from the long wait for the launch. “We didn’t really want to wait around for a year, but we had no choice.”
Engineers think that heating the valve may free it up. If that happens by mid-January, he said, LunaH-Map can maneuver into an alternative trajectory that would allow it to enter orbit around the moon in about a year.
Wide range of problems
The experience of LunaH-Map is emblematic of the challenges facing deep space smallsats. Spacecraft developers hoped the experience from building cubesats and other smallsats for Earth orbit could translate into more technically demanding missions to the moon and beyond. An early success was NASA’s twin Mars Cube One, or MarCO, cubesats that accompanied the InSight mission to Mars in 2018 and relayed telemetry from InSight as it landed.
More than half of the cubesats launched on Artemis 1, though, suffered problems after launch that, at a minimum, jeopardized their missions. The problems affected cubesats built by both space agencies and startups, and had little technically in common.
One of the most ambitious cubesats on Artemis 1 was OMOTENASHI, a contrived acronym for “Outstanding Moon Exploration Technologies Demonstrated by Nano Semi-Hard Impactor.” Developed by the Japanese space agency JAXA, OMOTENASHI was designed to land a small probe, weighing less than a kilogram, on the moon using a solid rocket motor and an airbag.
However, controllers struggled to establish communications with OMOTENASHI after separation as the spacecraft had problems with attitude control and power. They could not correct the problems in time to perform the maneuver needed for the landing.
NASA’s NEA Scout also had lofty goals: the 6U cubesat would deploy an 86-square-meter solar sail, allowing it to maneuver to a small near-Earth asteroid and collect high-resolution images. NEA Scout, though, failed to make contact with controllers after launch. Controllers even sent commands to NEA Scout to deploy its solar sail, hoping it might be seen by telescopes on the ground, but there was no evidence the sail deployed.
Cubesats not nearly as complex as NEA Scout and OMOTENASHI had problems as well. The NASA-funded Cubesat to Study Solar Particles, or CuSP, returned telemetry for an hour shortly after its deployment but has not been heard from since. Ominously, the last data from the spacecraft showed a temperature spike in one of its batteries.
Lunar IceCube, another NASA-funded cubesat to orbit the moon and look for water ice, made contact with controllers shortly after deployment. However, in a Nov. 29 update, NASA’s Goddard Space Flight Center said that the mission team was “continuing its attempts to communicate with the CubeSat so that it can be placed into its science orbit in the coming days.” NASA has not provided an update since then, and the mission’s principal investigator at Morehead State University did not respond to a request for comment.
Lockheed Martin’s LunIR spacecraft was designed to just fly by the moon, demonstrating an infrared sensor. However, the company said in December that an “unexpected issue with our radio signal” kept the spacecraft from conducting any observations during the flyby. Nonetheless, Lockheed called the mission a successful technology demonstration by proving that an infrared sensor and cryocooler could be packaged into a 6U cubesat.
At the opposite end of the business spectrum from Lockheed Martin is Miles Space, a startup whose Team Miles cubesat was also on Artemis 1. The company won a slot on the launch through a NASA Centennial Challenges competition to test propulsion and communications technology.
Wes Faler, chief executive of Miles Space, said that cubesat appears to be tumbling and is slightly off its planned trajectory. Ground stations are getting only fragments of individual data packets from the spacecraft as its transmitter goes in and out of view. “We’re working on software to assemble a data packet from the fragments, a process akin to sequencing a genome from DNA fragments,” he said in early January. “That’s going to take a while.”
SIMPLEx becomes complex
LunaH-Map was funded by NASA’s planetary science division through a program called Small Innovative Missions for Planetary Exploration, or SIMPLEx. It was one of two cubesat missions selected in 2015 for development.
The other original SIMPLEx mission was the CubeSat Particle Aggregation and Collision Experiment (Q-PACE), which would study how small particles collide and form larger particles in microgravity to help scientists understand the formation of solar systems. Q-PACE was placed into low Earth orbit on Virgin Orbit’s first successful LauncherOne mission in January 2021, but the cubesat never made contact with the ground.
Even before either of those missions launched, NASA moved on to a second round of the SIMPLEx program, focusing on larger smallsat missions. The agency selected for development three missions in 2019: the Janus mission to fly by binary asteroids, the Lunar Trailblazer lunar orbiter to search for water ice, and Escape and Plasma Acceleration and Dynamics Explorers (EscaPADE), a mission to study the interaction of the solar wind with Mars. Each mission had a $55 million cost cap.
All three suffered problems, some beyond their control. Original plans called for both EscaPADE and Janus to fly as secondary payloads on Psyche, a Discovery-class mission to the metallic main belt asteroid of the same name. A change in mission design for Psyche caused by switching from a Falcon 9 to a Falcon Heavy launch vehicle, though, meant that the mission could no longer drop off EscaPADE during a Mars flyby as originally planned.
EscaPADE has since redesigned its mission, using Photon satellite buses from Rocket Lab, and could launch as soon as 2024. However, NASA has yet to announce how it plans to launch EscaPADE.
Janus, meanwhile, is in limbo after software development problems with Psyche caused it to miss its August 2022 launch window. While Psyche has been rescheduled for launch in October 2023, that opportunity does not offer Janus any trajectories that would allow it to fly by binary asteroids of interest.
It’s unclear when or if Janus will fly. “I don’t have another ride for Janus,” Lori Glaze, director of NASA’s planetary science division, said at a town hall during the AGU Fall Meeting in December.
She added there were previously undisclosed issues with the propulsion system on Janus. “The certainty with which we could execute the mission is in question,” she said. She didn’t elaborate on the problems with Janus, and Lockheed Martin, which built the spacecraft, didn’t provide details about the issue.
Glaze said she’s allowing the team to use their remaining funding to look at alternative missions that could be performed with the Janus spacecraft, but offered no guarantees. “They could come back to us with a new proposed plan, but at this point it would be a new mission.”
Lunar Trailblazer had a very different problem: the lunar orbiter was, at one point, on track to be ready for launch at the end of 2022, but its original launch as a secondary payload on NASA’s IMAP space science mission had slipped to early 2025. In June 2022, NASA announced it secured an alternative ride for the spacecraft as a secondary payload on IM-2, the second commercial lunar lander mission by Intuitive Machines scheduled for launch as soon as mid-2023.
The spacecraft itself, though, faced cost overruns after the spacecraft’s prime contractor, Lockheed Martin, determined it required additional engineering and design work. NASA subjected the mission to a review and, in November, decided to proceed with it at a revised cost of $72 million, more than 30% over its original cost cap.
Despite the struggles with the SIMPLEx missions and the Artemis 1 cubesats, smallsats have had some successes beyond Earth orbit. A second JAXA cubesat on Artemis 1, EQUULEUS (for Equilibrium Lunar-Earth point 6U Spacecraft), successfully flew by the moon and tested a water-based propulsion system that placed the spacecraft on a low-energy trajectory to the Earth-moon L-2 Lagrange point. NASA’s BioSentinel cubesat was tumbling after deployment. Still, controllers were able to stabilize the spacecraft and maintain contact with it as it flew by the moon to study the effects of radiation on microorganisms.
ArgoMoon, a technology demonstration cubesat built by Italian company Argotec for the Italian space agency ASI, took images of the Earth and moon after deployment. However, the company said it needed more time than originally planned to commission the spacecraft. That success came after another Argotec-built cubesat, LICIACube, deployed from NASA’s Double Asteroid Redirection Test (DART) spacecraft and returned images of DART’s collision with Dimorphos, a moon orbiting the near-Earth asteroid Didymos, in September.
LICIACube exceeded expectations. “NASA was expecting to get a couple pictures” from LICIACube, said David Avino, chief executive of Argotec, in an interview in November. “We had 627 pictures taken by our spacecraft.”
He hopes the success of the two cubesat missions will generate more demand for its smallsats for both Earth orbit and deep space missions. “We want to have something that will not be cheap but will be reliable,” he said. “The main keyword is reliability. That means something that will allow our satellites to last up to five years, even in deep space.”
Even before launch, those who worked on the Artemis 1 cubesats knew that many of their satellites might malfunction. In an April 2022 report from a workshop on deep space smallsats, they compared their efforts to the early development of cubesats, when missions had high failure rates as designers struggled with the technical and cost limitations of such spacecraft.
Among the report’s recommendations was to shift from the 6U form factor used for the Artemis 1 cubesats to something larger, like 12U, to make it easier to accommodate components and dissipate heat. It also called for improvements in technologies for key subsystems, like attitude control and communications, and changes for how cubesats are accommodated as rideshares.
“Most of the reference development teams believe that many of the development issues discussed in this paper could be alleviated by using universally applicable solutions in ‘next generation’ cubesat mission architectures for deep space,” the report concluded.
That could include cubesats hitching rides on later Artemis missions. Jim Free, NASA associate administrator for exploration systems development, said during a webinar by the New York Space Business Roundtable Dec. 21 that NASA was preparing to fly cubesats on Artemis 2 and 3, the next two SLS launches.
Glaze said at the AGU Fall Meeting that, despite the challenges faced by some of its SIMPLEx missions, she would like to do more. The only difficulty is getting funding for them. “I love the SIMPLEx program, and I can’t wait to offer it again,” she said. “But I need some budget.”
Hardgrove, meanwhile, retained that cautious optimism about LunaH-Map he had before its launch. He noted in December that other spacecraft systems were working well, including testing its neutron spectrometer instrument as the cubesat flew past the moon. If its thruster is not fixed by mid-January, it may still be possible for the spacecraft to carry out an alternative mission, like an asteroid flyby, if it is restored later.
“We’re not dead. We’re doing great,” he said. “I think we’re hopefully going to ignite our propulsion system soon.”
This article originally appeared in the January 2023 issue of SpaceNews magazine.