Speaking to a St. Louis audience in September, Pam Melroy outlined the next space revolution, adding that it will happen in geosynchronous orbit. "What if you could build a satellite up there in GEO?" she asked. "What if you can repair it? What if you could upgrade it with the latest electronics?"

Perhaps fittingly, her talk at the Defense Advanced Projects Research Agency's "Wait? What?" symposium, was entitled "Port of Call at 36,000 KM."

Melroy, who as a NASA astronaut piloted two space shuttle missions, is intimately familiar with one of the foundations of DARPA's GEO Robotic Service (GRS) platform: a robotic arm. Both of her shuttle trips involved additions to the International Space Station (INS) in which its arm was used. The ISS arm also has been used in repairing the Hubble Telescope five times.

But that was at 347 miles aloft, in low Earth (LEO) orbit. GEO presents other issues, requiring greater capabilities. For example, the arm would have to include more automation and safety features. The last thing anybody wants is a malfunction that can knock out a satellite function and/or create debris in GEO.

"(DARPA's arm has) interesting characteristics, like robot reflexes and compliance control to greatly minimize the risk of debris from inadvertent collisions," said Melroy, now deputy director of the agency's Tactical Technology Office.

The platform would be part of an infrastructure, not just a one-stop shop for satellites in GEO orbit. "We think this is a critical capability toward building a transportation hub that allows transportation to and from the Earth's surface, from LEO orbit to GEO and even beyond the Earth's orbit," Melroy said. "We think that these capabilities are not just about a single monolithic satellite with a few capabilities, but instead are about a vibrant, robust ecosystem that involves transportation, repair, refueling, upgrading, in situ construction.

"Look at the great seafaring port cities in the world for inspiration and imagine a port of call at 36,000 kilometers."

The majority of satellites operating in geosynchronous orbit are commercially owned and operated, and so last year DARPA called upon companies like Intelsat General - which has done research on the idea for many years - for technical help and security advice in a Request for Information. The RFI sought insight into such things as potential tools that would be needed on a robotic arm in future GEO service. DARPA also wanted to know what modifications might be needed in existing orbiting satellites to make them receptive to servicing.

"We're asking the space community to think hard about how they want the future of space operations and how GEO robotics could help," said Gordon Roesler, a DARPA program manager specializing in space robotics, in a news release. "Their insights are essential as we take the first concrete steps toward viable satellite-servicing capabilities in GEO. If we're successful, we will significantly accelerate development of a capacity to maximize the utility of current space infrastructure and enhance the capabilities of future systems."

The agency also seeks advice in formulating a business case that would make the GRS platform financially feasible. DARPA has talked of a "fee-for-service" arrangement with military and commercial users.

"Creating a public-private partnership is an innovative way to ensure that GEO robotic servicing gets community buy-in to succeed long term," Roesler said. "Strategic partnership with commercial firms will be pursued in areas that both stabilize costs and improve the resilience of space architectures on which we rely."

The RFI brought massive input from the commercial community, including from Intelsat General, according to DARPA.

The robotic platform is one of three concepts that comprise the DARPA Phoenix program.

Another is a "satlets" idea in which production-line-possible satellite components weighing about 15 pounds (7 kilograms) each could be assembled in space. Looking much like Legos, the components encapsulate such satellite aspects as power, sensors and thermal management capacity. They could be assembled in different ways to accommodate different missions.

Another Phoenix approach is a Payload Orbital Delivery (POD) system that would standardize satellites and/or their components to take advantage of hosted payload opportunities offered by commercial satellites. The payloads would be about 1.3 feet by 1.6 feet by 2.2 feet and weigh between approximately 150 and 220 pounds, and would use a standard interface to attach the POD to the host satellite and release it at the proper orbit.

Those concepts, combined with the GRS platform can become a new approach to geosynchronous orbit satellites, both military and commercial. They offer opportunities to change orbits while in GEO, to inspect satellites as part of routine maintenance, to troubleshoot problems and repair components or replace those that are worn out, even to salvage parts off satellites that have completed their lifespans. Capabilities can be added or upgraded.

It could achieve the long-standing aim of replicating a terrestrial paradigm of "assemble, repair, upgrade, reuse," but doing so at GEO orbit 36,000 kilometers over Earth.

See a video of Melroy's presentation