From Sketch Pad to Launch Pad: U.S. Naval Research Laboratory’s Space Robotics Revolution

Over two decades ago, a group of U.S. Naval Research Laboratory (NRL) scientists, engineers, and researchers put pen to paper, brainstorming what would eventually become the first enduring robotic satellite servicing mission.

Funded by the Defense Advanced Research Projects Agency (DARPA), NRL designed and developed the Robotic Servicing of Geosynchronous Satellites (RSGS) payload. In late 2024, DARPA delivered the payload to SpaceLogistics, a Northrop Grumman company, who integrated it with their spacecraft bus, to become the Mission Robotic Vehicle (MRV), which is scheduled to launch at the end of this month.

NRL’s space capabilities focus on securing and defending national security and pushing the boundaries of space science for naval and Department of War (DoW) applications.

NRL’s Naval Center for Space Technology (NCST) supports the full scope of space operations—from ground systems to the launch pad and beyond.

“To put it simply, NRL goes everywhere in space,” said Bernard Kelm, Acting Director of NCST. “This journey to the launch pad represents the culmination of a multiple decades-long endeavor of vision, risk, and relentless engineering.”

A $2 Million Gamble and the Birth of a Lab In the late 1990s, many experts considered executing autonomous rendezvous and docking between two uncrewed spacecraft to be pure science fiction. Yet, forward-thinking researchers at NRL envisioned a future where the Department of Defense [now known as the Department of War] would need an orbital servicing vehicle not only to support satellite maintenance and upgrade capabilities but to also enable a new paradigm with even greater in-space capabilities than would otherwise be possible.

Convinced of this future, NRL Branch Head Sam Hollander (ret.) and NRL leadership authorized a $2 million Capital Improvement Program investment to build a space robotics rendezvous and proximity operations research laboratory. The risky investment functioned as a loan against NCST’s future. The state-of-the-art robotics facility completed construction in 2001 and was initially used for internally funded projects involving the deployment of large truss structures.

That initial bet paid off, enabling a broad spectrum of scientific research and development (R&D) and attracting many times its initial investment in external funding. By 2002, the facility captured the attention of DARPA program manager Gordon Roesler (ret.). He funded NRL to execute a small-scale study called RescueSat, exploring how to save billion-dollar satellites stuck in the wrong orbit. This study quickly evolved into the SUMO (Spacecraft for the Universal Modification of Orbits) program to begin maturation of the critical technologies for such a mission.

The SUMO Challenge & The “Make or Break” Demo “The SUMO mandate was daunting,” said Glen Henshaw, Ph.D., NRL Lead Space Roboticist for RSGS. “DARPA challenged us to design a robot that could dock with any satellite in space.”

Because satellites are custom-built without universal design standards, the team of engineers first ruled out nets, harpoons and tractor beams.

“We then realized a universal truth—every satellite got to space on a rocket!” he continued. “By targeting the sturdy ‘launch vehicle interface plane’—the structural ring or explosive bolt holes that attach a spacecraft to a rocket for launch—we determined that a robotic arm could safely grapple almost any spacecraft without damaging delicate instruments.”

NRL procured laboratory-grade equipment and launched into testing. The team encountered many technical challenges: the lighting in space is either blindingly bright or pitch black, and computer vision algorithms struggled to read the wrinkled features of thermal blankets. The NRL team met this challenge head-on by engineering laboratory test environments to simulate, as faithfully as possible, in-space conditions, including harsh orbital lighting.

DARPA laid down a clear goal for NRL: successfully grapple a mockup on two types of satellites and prove reliability. With funding dwindling and the program’s future on the line, the team worked tirelessly. The system teetered on the edge of success, with their autonomous robotic control algorithms acing two out of every three attempts. Knowing they might get only one shot to prove their concept, the engineers meticulously filmed every test. On the day of the sponsor’s visit and official demonstration, their persistence paid off. Their demonstration combining prototype robotics and software worked all three times.

The successful demonstration video was taken straight to the DARPA director, who greenlit the next phase of scientific R&D.

“The program is half-invisible,” Kelm said. “The flight software and robotic control algorithms are often overlooked but are just as important and just complex as the robotics hardware. NRL’s in-house expertise developed the enabling algorithms and software that led to the autonomy of RSGS.”

FREND, Mars Rovers, and the Human-Robot Teaming “Under the SUMO program, we proved that the mission was entirely doable,” Henshaw said. “But we also learned important lessons about how robotic control algorithms perform in space. Real flight microprocessors are built for extreme durability—much slower than the processor on an everyday laptop. Running the software under those realistic, restricted conditions forced us to go back to the drawing board and redesign our algorithms to be faster and more efficient.”

By 2005, the program evolved into FREND (Front End Robotics Enabling Near-term Demonstration). With funding in place, FREND aimed to build actual spaceflight robotic arms while advancing the flight software and other supporting technologies. Next, the team moved to secure vendors capable of manufacturing these arms.

NRL and DARPA embarked on a whirlwind procurement tour, visiting five serious bidders across North America in a single week, ultimately selecting Alliance Spacesystems—the same team that built the robotic arms for NASA’s Mars Curiosity Rover. The company delivered an Engineering Development Unit, which is still in use in NRL's Space Robotics Laboratory, and a Flight Prototype Unit. In 2008, NRL successfully subjected the protype to spaceflight environmental testing.

Despite these breakthroughs, finding a sustainable business case for the technology proved challenging. In response, the team shifted their focus to detailed application studies and continued low level development of critical technologies.

The largest application study was the Manned GEO Servicing study, a joint NASA-DARPA effort directed by the presidential administration to advance civil and national security collaboration, with a goal to explore human-robot-servicing of satellites in very high-altitude orbits. With an all-star team of defense and civilian space leaders, including NASA astronaut Piers Sellers and International Space Station flight controllers, the study solidified the value proposition of autonomous robotics. In 2012 DARPA transitioned the technology back into research and hardware maturation via a spaceflight concept program called Phoenix.

Systems Engineering, Myth-Busting, and a "Night at the Museum" Throughout development, skeptics claimed robotic servicing was impossible. Critics argued that satellites might tumble too fast, they might float away at first contact, that processing power was too weak, or that static discharge between two approaching spacecraft would cause catastrophic electrical failures.

NRL’s systems engineer systematically dismantled these myths through rigorous testing.

To test the theory that static discharge would cause catastrophic failures, NRL's Plasma Physics Division simulated the orbital environment. Simultaneously, the team conducted extensive air bearing and Electromagnetic Interference (EMI) chamber tests. These trials proved the robotic arms could safely secure a client satellite without causing it to drift or disrupting its antennas. Ultimately, the experiments demonstrated that the system thrived in zero-gravity and that any static discharge remained negligible and easily controlled.

“This testing was a substantial portion of the work NRL did under the FREND program,” Henshaw said. “In 2007 and 2008, the improved algorithm stack and the FREND arms came together for a series of demos for DARPA. Those demos proved to be the single leading milestone on our path to spaceflight, paving the way to the kickoff of the RSGS program.”

NRL engineers also solved incredibly complex software-hardware interactions. For instance, as a robotic arm moves, the lights and cameras mounted on it move as well. This causes shadows on the target satellite to shift, requiring the machine vision to constantly update in real-time. Furthermore, the physical movement of the robotic arm can induce motion in the satellite's solar arrays, so the team developed flight software that could autonomously correct for motion in a satellite's solar arrays caused by the movement of the robotic arm.

To test relative navigation sensors against large, fine-mesh satellite reflectors, the team secured exclusive after-hours access to the Smithsonian's Udvar-Hazy Center. For two consecutive nights, from 5:00 PM to 5:00 AM, NRL engineers rigged and operated millions of dollars in advanced spaceflight cameras and sensors in the museum’s dark hangar, suspending them around and above the Tracking and Data Relay Satellite (TDRS) displayed over the Space Shuttle Discovery. This "Night at the Museum" yielded critical data on how robotic sensors perceive complex structures in the harsh lighting of space. Particularly with the Smithsonian’s TDRS having spacecraft authentic components for the team to image – hardware that was not readily available elsewhere.

Forging the Future with RSGS In 2015, DARPA decided to move forward with what is now known as the RSGS payload. The decision established the core mission areas for the RSGS program: ultra-close inspection, orbital "tow truck" relocation, mechanical anomaly repair, and upgrading satellites not originally designed to be upgraded.

In 2019, DARPA executed a competitive procurement for a commercial partner to join the team and provide an enduring servicing capability. They selected SpaceLogistics, a Northrop Grumman company, to join the public-private partnership, combining the government-funded RSGS payload and SpaceLogistics commercially funded spacecraft, known as the MRV. Once the vehicle clears the launchpad, DARPA will transfer ownership of the furnished payload to Northrop Grumman. Following the completion of initial in-orbit capability demonstrations, DARPA intends to transfer the U.S. governments support of the RSGS program to the U.S. Space Force to support its Servicing, Mobility, and Logistics portfolio throughout the MRV's expected 10+ year operational life.

Operating 22,000 miles above Earth, geosynchronous orbit (GEO) is what Kelm calls the "Boardwalk and Park Place of Space.” It is highly valuable real estate critical for national security monitoring, military communications, and countless commercial operations. However, it is also becoming crowded with decommissioned unmanned spacecraft.

“Currently, industry designs satellites to be completely reliable, with zero room for error or unexpected failure, over their entire lifespan,” Henshaw said. “Think of it like a car. If you had to build a vehicle to drive 1 million miles without a single breakdown, it would cost an astronomical fortune. But we don’t do that because we can easily take our cars to a mechanic. Because that safety net exists, our vehicles don’t need to be over-engineered to last forever.

Satellites are so expensive because they lack that safety net; they must be built to survive that grueling ‘1-million-mile’ journey alone. But if we had a mechanic in space, we could build satellites much more affordably, without the need for such extreme, hyper-exquisite engineering.”

This public-private partnership between DARPA and SpaceLogistics intends to make routine on-orbit intervention a reality. The MRV, equipped with dexterous robotic arms, can perform tasks once thought impossible. It is designed to safely capture and repair older satellites never designed to be handled. It can perform orbital adjustments, latching onto a client satellite to install a propulsion "jetpack", developed by SpaceLogistics as the Mission Extension Pods, to extend its life. With the ability to carry interchangeable tools for in-flight resupply, it is built to resolve unforeseen mechanical anomalies on the fly.

The RSGS system is built for long-term endurance; it is designed to be resupplied in orbit via specialized cargo pods that deliver new tools and hardware as needed. This capability allows the robot to adapt to customer servicing missions not yet conceived. This adaptive approach is crucial because servicing legacy satellites, those of which were never designed to be touched by robotic hands, presents the most complex and challenging phase of on-orbit servicing .

By enabling these operations, the RSGS system not only enhances the resilience of current U.S. space infrastructure but serves as a precursor to future revolutionary capabilities, such as building large structures directly in orbit. As satellite servicing becomes more routine, future space vehicles will likely be designed with upgrades and servicing in mind, enabling unprecedented operational flexibility and cost savings.

From a sketch pad to a launch pad at Cape Canaveral, NRL has pioneered the next step in revolutionizing the space industry. The RSGS payload on MRV delivers an unprecedented and continuously evolving capability, limited only by the imagination of its users.

About the U.S. Naval Research Laboratory NRL is a scientific and engineering command dedicated to research that drives innovative advances for the U.S. Navy and Marine Corps from the seafloor to space and in the information domain. NRL is located in Washington, D.C. with major field sites in Stennis Space Center, Mississippi; Key West, Florida; Monterey, California.

NRL offers several mechanisms for collaborating with the broader scientific community, within and outside of the Federal government. These include Cooperative Research and Development Agreements (CRADAs), LP-CRADAs, Educational Partnership Agreements, agreements under the authority of 10 USC 4892, licensing agreements, FAR contracts, and other applicable agreements.

For more information, contact NRL Corporate Communications at mailto:NRLPAO@us.navy.mil.