Human-Robot Collaboration in Space Exploration ISS-Based Astronaut Controls Robot Team in Final Phase of ESA's Surface Avatar Experiment

Source: ESA 2 min Reading Time

Related Vendor

In a groundbreaking test of human-robot collaboration, an astronaut aboard the International Space Station remotely commanded multiple robots in a simulated Martian environment in Germany. The fourth and final phase of the ESA-DLR Surface Avatar experiment marks a key milestone toward future lunar and Martian missions.

In the final phase of ESA and DLR’s Surface Avatar experiment, astronaut Jonny Kim remotely commanded robots from the ISS, including ESA’s Spot and DLR’s humanoid Rollin’ Justin, in a simulated Martian environment.(Bild:  ESA Standard Licence)
In the final phase of ESA and DLR’s Surface Avatar experiment, astronaut Jonny Kim remotely commanded robots from the ISS, including ESA’s Spot and DLR’s humanoid Rollin’ Justin, in a simulated Martian environment.
(Bild: ESA Standard Licence)

This summer, the European Space Agency (ESA) and the German Aerospace Center (DLR) completed the final phase of their Surface Avatar project—a multi-year campaign designed to test how astronauts in orbit can command robotic systems on planetary surfaces. The simulation took place in Germany and featured a complex set of tasks executed by an advanced robot team under remote control from space.

NASA astronaut Jonny Kim operated the robots from the ISS using a specially developed control interface combining a joystick and haptic feedback. The user interface allowed switching between immersive first-person views and top-down mission maps, offering flexible and efficient control over multiple autonomous units.

In this fourth phase, the experiment introduced new levels of complexity and autonomy. Kim remotely controlled two robots: Spot, a four-legged robot from ESA capable of autonomous navigation, and Rollin’ Justin, a humanoid robot developed by DLR. Their task was to locate and collect scattered sample containers and deliver them to a mock lander. Spot navigated the terrain autonomously, while Rollin’ Justin was operated via a combination of direct control and pre-programmed instructions—freeing the astronaut to focus on high-level mission decisions. This marked a clear advancement from previous test phases that relied entirely on manual teleoperation.

A second test scenario demonstrated adaptive learning in robotic systems. ESA’s Interact Rover transported DLR’s quadruped robot Bert to a simulated cave entrance. After Kim cleared a boulder, Bert was deployed to explore the cave but simulated a leg malfunction mid-mission. The astronaut had to retrain Bert's locomotion algorithm in real time to allow the robot to continue searching for signs of Martian ice. This scenario aimed to evaluate the astronaut’s ability to handle unexpected issues and adapt robot behavior on the fly—critical skills for upcoming deep space missions.

Over the course of four test campaigns, the Surface Avatar team has refined methods of human-robot interaction, including task delegation and control handoffs between astronauts and autonomous systems. The results offer vital insights into which tasks astronauts prefer to retain and which can be reliably outsourced to robotic assistants—an essential distinction for future long-duration missions to the Moon and Mars.

Subscribe to the newsletter now

Don't Miss out on Our Best Content

By clicking on „Subscribe to Newsletter“ I agree to the processing and use of my data according to the consent form (please expand for details) and accept the Terms of Use. For more information, please see our Privacy Policy. The consent declaration relates, among other things, to the sending of editorial newsletters by email and to data matching for marketing purposes with selected advertising partners (e.g., LinkedIn, Google, Meta)

Unfold for details of your consent