WASHINGTON: Imagine self-healing satellites built in space. One sensor breaks down and another sensor elsewhere on the satellite takes up the slack. And the satellites are launched in modular pieces, on a series of different rockets, then are assembled by a robot arm in orbit. Parts can be replaced. The satellite can be refueled to prolong its operational life. This isn’t yet the space dock servicing the USS Enterprise, but it may mark the first steps toward such a capability.

It’s the stuff of science fiction and, as often happens, the Pentagon’s Defense Advanced Research Program Agency (DARPA) is the place experimenting with the new technologies and concepts of operations to make it real. Since the program, known as Phoenix, is expected to consume a tiny $40 million to $50 million this year (in its third year), don’t expect miracles any time soon.

Today, satellites have to be built with the knowledge that once they go into orbit they’re on their own. If something goes wrong or they just get old and die, there’s little that can be done beyond sending corrective signals to them from Earth and hoping they fix the problem. Phoenix would change that equation.

Part of the approach, Phoenix program manager David Barnhart says, relies on what he calls “satlets,” small, self-contained pieces of satellites that would be plugged together to create the larger birds. The original Phoenix concept, as its name implies, was to bring old satellites back to life by refueling them or sticking new sensors or other gear such as communications transponders on them.

The Phoniex program has already built a prototype robot arm to help refuel, repair, or build those satellites and has created a modular payload box to take the components into orbit.

But the technology clearly can be used for a wider array of purposes. One of the other benefits of this approach would be that the various components could be launched as what are known as hosted payloads. In other words, they hitch a ride with a larger satellite. Hosted payloads have long been advocated as one way to bring down the sky-high costs of launching satellites. But while commercial satellite firms hoped military programs would flock to them and bring down their own costs, the Pentagon has worried about a whole constellation of potential problems: control of their sensors should they be launched with a commercial satellite; interference from other sensors or transponders on the satellites; and quality control when the military payload is mated with the host.

Harris and Iridium came up last year with an intriguing approach, building a satellite with built-in space for hosted payloads. And the Air Force successfully deployed a sensor called CHIRP (Commercially Hosted Infrared Payload) that can serve as the beginning of an array of staring sensors designed to look for missile launches, other explosions and provide some change detection data. CHIRP, designed and built by Leidos (formerly Science Applications International Corporation), lasted three times as long as planned and was, according to the Air Force, a complete success. But resistance to the concept remains deeply embedded in Space and Missile Systems Center, the folks who build military satellites. Several civilian and defense space experts I spoke with yesterday at the industry conference Satellite 2014 said the culture change, underway for at least five or six years, remains a persistent problem.

The challenge was put best by someone who has wrestled with this issue for more than a decade at the British company, Surrey Satellite Technology Ltd. a leader in quickly building small satellites.

“Gen.  [William] Shelton [who heads Air Force Space Command] says he wants to do more of this, but what he has behind him is a large institutional body I call a supertanker, and it takes a long time to turn it around,” John Paffett, the CEO of Surrey’s American subsidiary, said yesterday during a panel on hosted payloads.

Combine the cultural barriers with the technological barriers and it is likely to be a long time before either DARPA’s vision of building and repairing satellites or launching many sensors into space as hosted payloads becomes a reality.