Space robotics

Robots do the work in space that crews cannot afford to do and cannot reach. The engineering is shaped by three constraints absent on Earth: no gravity to react against, no opportunity for repair, and a control loop that may take minutes to close.

Manipulators

A robotic arm on Earth pushes against a floor bolted to a planet. In orbit, the arm and the spacecraft form one dynamical system, and moving the arm moves the spacecraft.

The flown solutions are large and slow. Canadarm and Canadarm2 move deliberately, with the station’s attitude control coordinated against planned arm motion. Canadarm2 can also relocate itself end over end between grapple fixtures on the station’s exterior, effectively walking, which extends its reach far beyond its length.

The Mission Extension Vehicle demonstrated the commercial case for manipulation without actually manipulating much. Rather than repairing or refuelling a satellite, it docks to the target’s apogee engine nozzle, a structure present on nearly every geostationary satellite, and then provides attitude control and station keeping using its own propulsion. It is a deliberately minimal approach that avoids the hard problems, and it works. See In-space manufacturing and assembly.

Surface rovers

Mars rover driving is constrained by the communication cycle rather than by speed. A traverse is planned on the ground from the previous day’s imagery, uplinked, and executed. The rover then waits for the next cycle. Even with autonomous navigation, which lets a rover assess terrain and pick a path itself, throughput is limited to roughly one command cycle per sol.

Autonomous navigation builds a terrain model from stereo imagery, evaluates traversability, and plans a path. Its value is measured directly in metres per sol: it lets a rover drive beyond what the operators could see when they planned, which multiplies daily progress.

Slip is the persistent problem. Wheels turn without the rover moving on loose regolith, which is why visual odometry is essential rather than supplementary. See Computer vision in space.

Wheel design has proved harder than expected. Curiosity’s thin aluminium wheels sustained significant damage from sharp embedded rocks earlier than predicted, requiring route planning to avoid certain terrain and a change in driving technique. Perseverance’s wheels were redesigned as a direct result.

Free flyers

Small robots that fly inside or around a spacecraft.

Astrobee and its predecessor SPHERES operate inside the ISS, using fans for propulsion and vision for navigation. Their value is routine work: inventory audits, camera work and sensor surveys that would otherwise consume crew time, which is the scarcest resource on a station.

External inspection free flyers can survey a spacecraft’s exterior for damage, which is otherwise only possible by EVA or by another spacecraft. The difficulty is that anything flying near a spacecraft is a collision risk, so proximity operations require the same care described in Rendezvous and docking.

Teleoperation under delay

Direct teleoperation degrades badly with delay. Above roughly a second of round trip, an operator controlling a manipulator directly will overshoot and oscillate, because the feedback arrives after the correction has been applied.

Predictive display helps in the intermediate range by showing the operator a simulated result of their command immediately, with the actual telemetry arriving later to correct the model. It works when the model is good, which means when the environment is well characterised.

At Mars delays, direct control is impossible and the only workable mode is supervised autonomy: the ground specifies a goal and the robot achieves it, with the ground reviewing the result a cycle later. This makes robot capability the throughput limit, which is why rover autonomy translates so directly into science return.

The unsolved problem

Grappling an uncooperative target remains beyond operational capability, and it is the enabling technology for both active debris removal and servicing the existing satellite fleet.

Approaches under development include robotic arms with vision-guided grasping, nets, harpoons, and adhesive or electrostatic capture. Each has failure modes that risk generating debris, which is the outcome the mission exists to prevent.

Demonstrations have made real progress. RemoveDEBRIS tested net and harpoon capture on deployed targets, Astroscale’s ELSA-d demonstrated magnetic capture of a prepared client, and ADRAS-J performed close approach and inspection of an actual derelict upper stage. All are proving technology rather than operating a service. See Space debris and traffic management.

Design constraints

Lubrication is the underappreciated one. Liquid lubricants evaporate and creep in vacuum, so space mechanisms use low-vapour-pressure oils in sealed or labyrinth arrangements, or solid lubricants such as molybdenum disulfide. Getting this wrong produces a mechanism that works perfectly in test and seizes in flight. See Structures and mechanisms.

References

• Flores-Abad, A. et al. “A review of space robotics technologies for on-orbit servicing”, Progress in Aerospace Sciences, 68:1, 2014.
• Maimone, M., Cheng, Y. and Matthies, L. “Two years of visual odometry on the Mars Exploration Rovers”, Journal of Field Robotics, 24:169, 2007.
• Rekleitis, I. et al. “Autonomous capture of a tumbling satellite”, Journal of Field Robotics, 24:275, 2007.
• Smith, T. et al. “Astrobee: A New Platform for Free-Flying Robotics on the International Space Station”, i-SAIRAS 2016.
• Forshaw, J. L. et al. “RemoveDEBRIS: An in-orbit active debris removal demonstration mission”, Acta Astronautica, 127:448, 2016.