National-security space programs have become so slow and costly that the U.S. faces the “self-inflicted surprise” of other nations being able to put capabilities into orbit much faster, says the Defense Advanced Research Projects Agency (Darpa) director.

Formed in 1958 to prevent a repeat of the “technological surprise” of the Soviet Union's Sputnik satellite launch, the Pentagon agency has embarked on a group of projects intended to make access to space quicker and cheaper. It is an effort to counter what Darpa Director Arati Prabhakar sees as a troubling development “to do with how slow and costly it is for us to do anything we need to do on orbit for national security purposes.”

The projects involve lower-cost, more-responsive launch systems for smaller satellites; building blocks that allow spacecraft to be produced more rapidly and cheaply; robotic technology to assemble, upgrade or repurpose satellites in orbit; and sensors to see and control what is going on in space.

Speaking at the American Institute of Aeronautics and Astronautics' SciTech 2014 conference here last week, Prabhakar said that as commercial and foreign space activity increases, the U.S. is “freezing in place . . . in terms of our ability to react and do what we need to do quickly, cost-effectively in space.” This is putting at risk the precise, lethal warfare capability “that is a core element of our national security today . . . [but is] simply not possible without the assets we have in space,” she says.

Over the past two decades, U.S. launches have declined and shifted to larger and heavier government satellites, and prices have increased, says Antonio Elias, chief technical officer for Orbital Sciences Corp. At the same time, international launches have increased, he told the conference.

Darpa's involvement in space has waxed and waned over the years, and it has targeted smaller, faster, cheaper satellite launches in the past. Agency funding helped Orbital Sciences develop the air-launched Pegasus, and Space Exploration Technologies develop the Falcon. But programs like Rascal, to develop an aircraft-based small-satellite launcher and System F6 to “fractionate” satellites into clusters of smaller formation-flying spacecraft, were terminated before hardware was flown.

But Darpa is always willing to try again. It has initiated the Airborne Launch Assist Space Access (Alasa) project to air-launch 100-lb. satellites into low Earth orbit for $1 million; the XS-1 reusable spaceplane to launch 3,000-5,000-lb. payloads for $5 million; and the Phoenix program to reuse non-working satellites in orbit by robotically attaching building-block “satlets.” Darpa believes such technologies can both unlock the commercial market for small satellites and finally persuade the Pentagon to “disaggregate” its big, expensive satellites into fleets of smaller craft that can be launched more quickly for less money.

Boeing was awarded the Alasa demonstration contract late last year, and flights are planned for 2015, says Brad Tousley, Darpa Tactical Technology Office director. Boeing will use the F-15 fighter to launch a rocket-powered second stage “using a new propulsion approach that will greatly reduce cost,” he says. The goal is to launch 100-lb. payloads from conventional runways with just 24-hr. call-up.

The XS-1 program aims to develop a reusable first stage that enables aircraft-like operations from a clean pad. Darpa wants to demonstrate 10 flights in 10 days; but the plan has its critics. It is a mistake to try to treat launch vehicles like aircraft, Elias says, adding, “[Darpa] is doing yet another study of a reusable launch vehicle—I can't believe it.”

The Phoenix program's concept of constructing satellites from building-block “satlets” could allow a spacecraft to be assembled quickly around the payload for a specific mission or service, cutting cost and speeding delivery, says its developer, NovaWurks. Each individual satlet has all the functions of a large satellite, but on a miniature scale, says Talbot Jaeger, NovaWurks founder. Instead of integrating the payload into a large satellite bus that provides the required mechanical, propulsion, power, thermal, communications, processing and other functions, the spacecraft—a conformal satellite—would be assembled around the payload by connecting multiple satlets together. “We reverse the standard engineering process. We do not have a bus to put the payload on; we form the bus to meet the needs of the payload,” Jaeger says.

The highly integrated satlet (HiSat) design is being developed for the Phoenix program to demonstrate that retired satellites can be repurposed in orbit. In October, NovaWurks was awarded a potential $40 million contract for Phases 2 and 3 of the program.

In Phase 1, the company completed a proof-of-principle demonstration of its satlet. “Phase 2 will take us to a viable prototype,” Jaeger says, adding the company will build a large number of satlet cells for testing. “Version 0” of the box-shaped HiSat measures 20 X 20 X 10 cm (8 X 8 X 4 in.).

If exercised, Phase 3 will be a flight demonstration in which the satlets are carried into orbit on a host satellite and released to be collected by a robotic servicer/tender spacecraft. The robot will attach the satlets to an antenna detached from a defunct satellite to reconstitute an operational spacecraft.

David Barnhart, Darpa's Phoenix program manager, says the agency plans a satlet flight test in 2015, under Phase 2, to prove they are robust and reliable enough to work in orbit and to pave the way for the robotic flight demonstration.

There are two architecture approaches to what Darpa calls satellite “cellularization”—a reference to the cellphone-like low-cost manufacturing paradigm. The heterogenous approach divides up the building blocks by function, with different satlets for propulsion, power, etc. The homogenous one aggregrates multiple identical satlets, each with all the functions.

“The HiSat has a portion of a satellite's functionality. As you build together more satlets, you build more functionality back into the spacecraft,” Jaeger says. “You need more than one to bring back a satellite's functionality,” he adds. While NovaWurks is pursuing a homogenous architecture, Darpa also has awarded a contract to Busek to demonstrate a heterogenous approach by developing a satlet dedicated to propulsion, Barnhart says.

Key to NovaWurks' architecture is a connector design that allows satlets to share fuel, power, thermal, data and other services so their functionality can be aggregated. There are connector pairs on each of the satlet's edges, and on a central turntable, so they can be assembled in different configurations. The connectors allow satlets to be joined—manually on the ground or robotically in space—and rotated into position around the payload or repurposed antenna to create a functioning spacecraft.

Compared with an aggregation architecture in which each building block has a different function, such as propulsion or communication, Jaeger says the homogenous satlets approach provides greater resilience to failures and flexibility to configure—or reconfigure—a satellite. The satlets “would be built not in tens or hundreds, but in thousands,” he says, to lower costs.