- Sarah Downs developed force-based algorithm for satellite antenna insertion
- System works without cameras in harsh space environments
- Downs now researching larger-scale satellite manipulation at Texas A&M
- NASA OSAM-1 mission faces $2.05 billion budget pressure
Sarah Downs built a robot that can insert a satellite antenna without using cameras. The University of Tulsa graduate student developed the force-based algorithm during her master’s thesis in collaboration with NASA and the U.S. Air Force, addressing the classic peg-in-hole problem that has challenged robotics engineers for decades.
The system deploys a torque sensor on the gripper to feel force feedback, sensing the position and orientation of objects without vision systems. While cameras typically guide satellite assembly work, they can malfunction or encounter delays in the harsh, remote environment of outer space.
Force sensing replaces vision for contact-rich tasks
Peg-in-hole assembly for large and brittle satellite workpieces poses a major challenge under contact-unknown conditions in smart manufacturing. For contact-rich manipulation tasks, force-torque sensing separates robots that can handle delicate assembly from robots that break everything they touch. Six-axis force torque sensors detect three force components and three torque components.
For robotic assembly, the required fit for insertion is typically too precise for visual feedback alone. Multiple peg-in-hole assembly technology spans tasks from on-orbit assembly of large space structures to precise assembly of small electronic components.
Downs’ work arrives as NASA pushes robotic satellite servicing. The agency’s OSAM-1 (On-orbit Servicing, Assembly, and Manufacturing) project faces pressure to stay within its $2.05 billion budget and December 2026 launch date due to poor contractor performance and technology development issues. OSAM-1 is planned as the first mission to robotically refuel a spacecraft not designed for on-orbit servicing, rendezvous with Landsat 7, and replenish its hydrazine fuel tank.
Middle school robotics led to aerospace research
Downs joined a Tulsa middle school First Lego League robotics team from 2014 to 2016. After her father died from a heart attack in 2015 when she was 13, her mother returned to college to earn a business degree. The financial pressure shaped Downs’ career thinking. She joined First robotics in high school and spent her final two high school years splitting days between regular classes and engineering courses at Tulsa Tech vocational school.
For her senior capstone at the University of Tulsa, Downs and two classmates designed a lunar lander exhibit for the Tulsa Air and Space Museum. The interactive game simulates missions on the moon, Venus, Mars, and Titan using a game controller. She earned her bachelor’s degree in electrical engineering in 2024.
When government funding delays postponed the NASA project, Downs spent her first graduate year at the university’s new Institute for Robotics and Autonomy. Inspired by her wheelchair-bound grandmother who had severe arthritis, she developed a robotic arm to help older people and wheelchair users live independently. The NASA project secured funding before her sophomore year in 2025.
Ph.D. research scales up satellite work
Downs is now a Ph.D. student in electrical engineering at Texas A&M University in College Station, continuing satellite assembly and manipulation research “on a much larger scale.” The progression mirrors broader industry shifts toward autonomous space infrastructure.
The ability to autonomously assemble structures is crucial for the development of future space infrastructure, though unpredictable space conditions pose significant challenges for robotic systems. On-orbit operations have become a research hotspot, but traditional on-orbit assembly using only robotic arms suffers from low precision.
The technical challenge extends beyond NASA’s immediate needs. While force sensing offers advantages in contact-rich scenarios, it complements rather than replaces vision systems in most terrestrial applications. Vision excels at general positioning and picking; force-torque sensors handle the final alignment and insertion where tight tolerances make visual feedback insufficient. The space environment simply removes the vision option, forcing engineers to rely entirely on tactile feedback—a constraint that could yield techniques applicable to terrestrial robotic assembly where camera occlusion or lighting conditions limit optical approaches.
Force-based insertion without vision systems solves a real space problem but addresses a broader assembly challenge: what happens when you can’t see the final alignment. For engineers designing assembly cells with occlusion issues or difficult lighting, the tactile-first approach offers a pathway that doesn’t depend on perfect camera placement. The technique scales down to terrestrial applications where vision struggles with reflective surfaces, confined spaces, or parts that obscure each other during final insertion.
What clearance tolerances can force-torque sensors handle for peg-in-hole assembly?
Dual-arm robots using six degrees-of-freedom force-torque sensors have successfully assembled round and square pieces with 0.5 mm clearance. Six-axis sensors measure forces along three axes and torques around three axes, enabling the precision needed for tight-tolerance insertions without requiring high-precision fixturing.
Why use force sensing instead of cameras for satellite assembly in space?
Cameras typically guide satellite assembly work, but in the harsh, remote environment of outer space, cameras might malfunction or encounter delays. Force-torque sensors mounted on the gripper provide tactile feedback that doesn’t depend on lighting, line of sight, or electromagnetic interference that can affect optical systems during orbital operations.
Article Source: This Graduate Student Equips NASA’s Robots With Assembly Skills
