LOGAN, Utah — Despite concerns about reliability and access to launch vehicles, the small satellite industry expects the number of cubesats to continue to grow as they find new commercial and government applications.

In a presentation at the Conference on Small Satellites at Utah State University here Aug. 8, Bill Doncaster of SpaceWorks Engineering said his company was maintaining a forecast issued earlier this year that predicted about 200 satellites weighing between 1 and 50 kilograms would launch this year, a number that would break the record of 158 set in 2014. That vast majority of those satellites would be versions of cubesats weighing 10 kilograms or less.

SpaceWorks, in a similar forecast last year, forecast 163 such satellites would launch in 2015, but only 131 actually flew. “That was an anomaly based on available launch slots,” he said. He noted that both Antares and Falcon 9 rockets, which have launched many such spacecraft on cargo missions to the International Space Station, were recovering from launch failures. “The number of opportunities was somewhat limited.”

Doncaster reiterated that prediction even though, as of his presentation, only 53 satellites in that mass class had launched so far in 2016. “I believe that that 200 number is still very achievable,” he said.

That conclusion was based on two launches scheduled for late 2016, one of a Cygnus mission to the ISS designated OA-7, and a SpaceX Falcon 9 launch with a secondary payload adapter called SHERPA. “If everybody squeezes their launches in this year, 2016 will be the highest year on record for number of launches in this mass class.”

SHERPA, developed by Seattle-based Spaceflight, will carry 87 satellites, most of which will weigh less than 50 kilograms. In an Aug. 8 conference presentation, Jason Andrews, president and chief executive of Spaceflight, said that SHERPA will ship to the launch site at Vandenberg Air Force Base in California in October. The launch, he said, is currently planned for late October, with SHERPA flying as a secondary payload on the Falcon 9 launch of Formosat-5 remote sensing spacecraft.

The timing of the OA-7 mission is less certain. A NASA ISS flight plan chart, dated July 7 and presented at a July 28 meeting of the NASA Advisory Council, listed a Dec. 30 launch date for OA-7. That was before NASA and Orbital ATK announced Aug. 10 that the OA-5 Cygnus mission, previously scheduled for launch Aug. 22, would be delayed until no earlier than the second half of September. That delay could, in turn, push back OA-7 into 2017.

Despite the near-term uncertainty about upcoming launches, Doncaster said SpaceWorks expected launches of cubesats and other very small spacecraft to grow for the next several years. “SpaceWorks believes that, over the next five years, that there could be up to 3,000 micro and nanosatellites that need a launch,” he said. “We think that the market is going to continue to grow. It’s real.”

Driving that growth is remote sensing, as several companies plan constellations of cubesats and other small satellites to provide imagery and weather data. SpaceWorks projects that nearly three quarters of all satellites weighing 1 to 50 kilograms launched from 2016 through 2018 will serve remote sensing applications. The company also predicts that 70 percent of all such satellites launched in the same period, regardless of application, will be commercial.

Reliability issues

That increasing interest in cubesats comes despite nagging concerns about their reliability. In an Aug. 10 presentation at the conference, Martin Langer of the Institute of Astronautics at the Technical University of Munich presented an analysis of 178 cubesats launched through mid-2014. Of them, 18 percent were “dead on arrival”; that is, they failed to operate at all after launch. “That’s way too high,” he said.

Diagnosing cubesat failures can also be difficult. Langer said that one third of all cubesat failures in his database had no known cause because of a lack of telemetry to diagnose. That includes the dead-on-arrival cubesats and some others. For cubesats that fail after more than 30 days in orbit, though, unknown failures drop to 12 percent, with power systems, communications and on-board computers accounting for most failures.

Langer noted many of the cubesats included in his analysis were built by universities, often as either educational or experimental efforts that have higher failure rates. A separate study of university cubesats and other smallsats, presented by Michael Swartwout of St. Louis University at the conference Aug. 11, found that more than 40 percent of those satellites were either dead on arrival or failed early in their missions, although those failure rates were lower for so-called “flagship” universities that have more experience with cubesats.

Langer said that his data also did not include any of the nearly 150 cubesats placed in orbit by Planet, the San Francisco company that builds cubesats to collect Earth imagery, because of a lack of information about any failures they experienced in orbit.

The solution to improving reliability of cubesats, Langer said, was not through the use more traditional mission assurance techniques or higher reliability components, both of which can impose much higher costs on cubesat programs. He advocated instead for more system-level testing of spacecraft to uncover problems prior to launch.

He also recommended greater sharing of information about cubesat failures among spacecraft developers. “For cubesats, we have to accept that failure is an option. We have to learn from that,” he said. “We have to provide our lessons learned to the community.”

Jeff Foust writes about space policy, commercial space, and related topics for SpaceNews. He earned a Ph.D. in planetary sciences from the Massachusetts Institute of Technology and a bachelor’s degree with honors in geophysics and planetary science...