Reusing hardware already in orbit, rather than launching a new spacecraft, could dramatically cut the cost of providing military satellite communications, but requires a technology leap to enable robotic servicing in geosynchronous orbit and beyond.

Under the Phoenix demonstration planned for 2016, the U.S. Defense Advanced Research Projects Agency (Darpa) plans to show that a robotic vehicle can remove the antenna from a retired spacecraft in graveyard orbit and attach systems to it to rebuild a functioning communications satellite in geostationary orbit (GEO).

Phoenix aims to “increase the return on investment for Defense Department space missions by really lowering the cost,” says David Barnhart, Darpa program manager. The program centers on developing “a different way of building spacecraft, with some level on on-orbit assembly to add hardware and use what is already up there.”

Although antennas average only 2-3% of a communications satellite's mass, the cost of the spacecraft increases proportionally with aperture size, in turn driving the size of booster and cost of launch into GEO. “The dollars per specific mass is not that high for an antenna, but if you have to send up a large satellite on a large booster the cost is very high,” he says.

An example is the $350 million sticker price on NASA's TRDS-K data-relay satellite, with its two 4.9-meter-dia. (16-ft.) unfolding parabolic antennas, and the likely $200 million price tag for its launch in January by Atlas V. “The cost depends on the size of the antenna. With a large commercial aperture, say 18 meters, this architecture could provide a potential 10 times reduction,” he says.

While it can take up to 15 years for fuel reserves to be depleted and solar arrays degraded to the point where a satellite has to be retired, Darpa calculates the antenna structure could last more than a century. The agency believes the cost of providing satellite communications could be reduced dramatically by removing and repurposing an existing antenna by attaching “satlet” modules manufactured at high volume and low cost and launched cheaply and quickly as piggyback payloads on commercial satellites.

“If we cannot replace the appropriate functions and mass to control the aperture, the concept does not make sense,” says Barnhart. The plan is to achieve this through aggregation, sending up satlets that each perform a function, such as attitude control, momentum management or power generation, provided by the original satellite. “We will create an aggregate set of hardware. The question is how many of these very small things do we need, and can we aggregate them at low cost?”

In the Phoenix concept, a robotic servicer/tender would be launched into GEO. The satlets, along with tools for the servicer's robot arms, would then be packed into payload orbital delivery systems (PODS) and delivered to GEO as hosted payloads on commercial satellite launches. “The tempo at which the satlets are going up becomes critical,” says Barnhart. With 10-15 GEO commercial launches a year on average, “that's a one-a-month tempo on which we can take advantage of any excess mass as hosted hardware.”

The PODS would be ejected on command from the GEO satellite and collected by the servicer, which would store the satlets and tools on its toolbelt before heading to graveyard orbit to rendezvous with the donor spacecraft. There the servicer would attach the satlets to the antenna, which would then be severed from its satellite and towed to GEO to take up position as a reconstituted communications satellite.

Begun last July, the four-year, $180 million Phoenix program is making progress with several prototype systems undergoing laboratory testing. These include:

•The servicer/tender's robotic-arm grappler, end-effector gripper and adhesive grasping pads to hold the satellite.

•A “hyperdexterous” mutli-jointed robot arm to bring lights and cameras close to the work area.

•A tool to mechanically sever the boom carrying the antenna while minimizing debris generation.

•A system to make a “stem boom” for gravity stabilization of the separated antenna from flat carbon-fiber tape.

This month, Darpa will issue a solicitation for the remaining five of 12 technologies required for the Phoenix concept. These involve elements critical to the planned demonstration, including sensors and systems for “safe and responsible” rendezvous and proximity operations by the servicer, both while picking up the drifting PODS and working on the cooperative donor satellite. Darpa will also explore with commercial satellite operators ways to launch the PODS into GEO.

The Phoenix demonstration will test key aspects of the concept, but not a complete operational system. “In the first demonstration we will attempt to bring up the functions to repurpose an antenna,” says Barnhart. “For the demo, we will take the satlets up with us [on the servicer/tender launch].” The demonstration aims to repurpose a small 1-5-meter antenna and validate that radio-frequency communications can be restored via the rebuilt aperture in graveyard orbit. “For the demo, we will not bring the aperture to GEO,” he says.

Darpa has yet to select a donor satellite. “Of the 500 retired spacecraft, 140 have been identified with apertures that would be useful for the demo. We are continuing to work through a number of identified candidate assets.”