ESA’s “flattest floor in Europe”: how ORBIT in the Netherlands trains robots and humans for life without gravity
The European Space Agency (ESA) in January 2026 published the results of autumn testing at its Orbital Robotics Laboratory (ORL) at the Dutch ESTEC, where three student teams from Europe tested technologies for future space tasks: autonomous “crawling” of a robotic arm over structures, a neuroscience experiment with human subjects, and capturing objects with a gecko-inspired gripper.Testing was conducted on ORBIT, a platform that mimics weightlessness in two dimensions: a system of “floating” platforms glides over an ultra-flat floor, frictionless, thanks to air bearings. Such an environment allows control algorithms, robotic mechanisms, and human interaction with “microgravity” to be verified on Earth before expensive hardware goes into orbit, where errors can mean mission loss.
What is ORBIT and why “flatness” is crucial
ORBIT is the central infrastructure of the ORL within ESA’s technical center ESTEC in Noordwijk. ESA describes it as an ultra-flat floor with a surface area of about 43 square meters (roughly 4.8 by 9 meters), where the difference between the lowest and highest point is less than one millimeter. The technical description states that the total flatness deviation is less than 0.8 millimeters (with tolerance), and the maximum slope is less than 0.3 millimeters per meter, which is the standard required for reliable “free-floating” dynamics without the hidden influence of slope or irregularities on the platform's movement.
Platforms with air bearings are used on ORBIT: an air cushion creates a stable air gap, thinner than a human hair, so the platform moves with almost no friction, similar to an air hockey table. In practice, this means that an instrument, a robotic arm, or a seat with a human subject on the platform behaves as if in weightlessness, but limited to a plane: two translational and one rotational dimension. For motion verification, ORBIT is also equipped with an optical tracking system (VICON), which provides reference data on the position and velocity of objects in space, which is important for comparing algorithm behavior with the “truth” of measurements.
ESA uses such a laboratory for a wide range of activities in orbital robotics and guidance, navigation, and control (GNC), including research on active space debris removal, satellite servicing in orbit, and visions of assembling large structures in space. The key is that systems that must operate in orbit near other objects, with very low relative velocities and without ground support, can be realistically verified on Earth in a controlled environment.
A program that opens the doors of “real” space laboratories to students
Three projects came to the ORL through the ESA Academy Experiments Programme, an annual educational program intended for undergraduate, graduate, and doctoral students from eligible states. ESA emphasizes that the program guides teams through the entire project development path – from concept and design to operations and data analysis – with an emphasis on industrial engineering practices, project management, risk reduction, and funding strategies. The program also includes formal phases (from orientation and education to consultations, selection, campaign implementation, and publication of results).
In addition to the ORL, the program opens up to other platforms depending on the edition: from parabolic flights and drop-towers to centrifuges, experiments on the International Space Station, and missions like Space Rider. In the cycle that began in 2025, ESA directed part of the projects specifically to parabolic flights and the ORL, and after evaluation, five teams were chosen – three for the ORL and two for microgravity campaigns in an aircraft.
SKYWALKER: a robotic arm learning to “grab” and pull its own base
The Danish team SKYWALKER from Aalborg University came to the ORL with a focus on autonomous robotic arm movement in microgravity using reinforcement learning. The idea is simple to describe but demanding to execute: the robot needs to find a support point on a structure, grab onto it, and then “pull” its own floating base to move to a new position. Today, such “crawling” movements of robots over structures in space are usually pre-programmed; the goal, according to the team members' explanations, is to show that the system can learn and adapt, which would increase robot autonomy in tasks of assembling and maintaining large space structures in the long run.
According to ESA, SKYWALKER was selected as part of the program in February 2025, and the test campaign at the ORL was held in late September 2025. In the laboratory, the team placed the robotic arm on the MANTIS floating platform, connected the system to VICON, and then separately verified two algorithms. The first algorithm was intended to allow the arm to independently detect the gripping point and “anchor” itself, while the second – technically more demanding – was supposed to coordinate the movement of the arm and the base, so that the platform moves while the arm is hooked.
Team member Rasmus Kristiansen pointed out that the first part was focused on autonomous finding and grabbing of the support point, while the second part – pulling the base with the arm's movement – showed differences compared to computer simulations in a real laboratory environment. That “clash with reality” is one of the reasons why ESA emphasizes the value of ORBIT: simulations often assume ideal conditions, while in the laboratory there are sensor limitations, system delays, mechanical tolerances, and tiny differences in the interaction of the platform with the surface that can “distort” the expected outcome.
Despite problems with the second algorithm, ESA states that the team managed to demonstrate the full “crawling” sequence by combining the movement of the arm and the platform, and that the students identified the cause of the deviations and created a plan for refinement after returning to the university. ESA’s official review also mentions that the project continued to progress after the campaign, including a recent graduate taking on full-time work on the project, which speaks to the transfer of student work into a more serious research and development direction.
V-STARS: the first experiment with human subjects on ORBIT
While SKYWALKER was focused on robotic autonomy, the British team V-STARS (Birkbeck, University of London and University of Kent) opened ORBIT to neuroscience and human physiology. ESA states that this is the first ORBIT experiment to involve human subjects, with a focus on the link between the vestibular system (the part of the inner ear crucial for balance) and the perception of the vertical in microgravity conditions.
According to the team’s description, a total of 22 subjects participated in the experiment. A participant sits on a chair attached to a floating platform, wears VR goggles, and has small electrodes placed behind the ear. The platform is then slowly pushed over the ultra-flat floor in random directions, while the electrodes send mild signals that stimulate the vestibular system. In the VR view, the participant sees a simple line, and researchers check the perception of verticality by asking the subject if they see the line as completely vertical or slightly tilted.
The key concept upon which V-STARS builds is vestibular stochastic resonance: a phenomenon where controlled “noise” can increase the sensitivity of the sensory system. The team is researching whether such an approach can improve perception and potentially accelerate adaptation to microgravity, which in space missions is linked to crew safety and efficiency, especially in the initial days of stay in orbit.
ESA states in a separate publication that V-STARS was selected in February 2025 and that, before the campaign itself, the team had to obtain ethical approvals in the United Kingdom and authorization from ESA’s medical board. The campaign, according to ESA, lasted two weeks, and more than 20 participants were tested, after which the team returned to the universities to analyze the results.
For the ORL, this was an important organizational step forward: a laboratory traditionally associated with robotics and satellite capture/servicing systems had to introduce procedures appropriate for working with humans, from safety protocols to the integration of equipment like VR systems and vestibular stimulation. Team member Milena da Silva Baiao pointed out that integrating the experiment into a new environment was a challenge, but also a valuable experience for the students and the laboratory team.
GRASP: gecko-inspired “sticky” gripper for non-cooperative targets
The Italian team GRASP from Sapienza University of Rome came to the ORL with a problem that is becoming increasingly important in orbits around Earth: how to safely approach a “non-cooperative” object and capture it without creating uncontrolled momentum or repulsion. Such scenarios include space debris, but also satellites that need to be serviced, redirected, or prepared for controlled removal from orbit.
GRASP stands for Gecko Rendezvous Autonomous System and Pincher. The team, which according to ESA gathers 14 students of various profiles (from aerospace engineering and robotics to artificial intelligence), developed a planar robotic manipulator with “tentacles” and pads made of material that mimics gecko adhesion. ESA's description states that adhesion is based on microscopic structures that, under mild shear forces, achieve contact and develop Van der Waals forces, without the need for glue or vacuum grippers, and with the possibility of multiple activation and release.
In ESA’s main publication, team member Stefano De Gasperin explains that their experiment simulated a small spacecraft approaching an object and capturing it using a tentacled gripper with adhesive pads. Lorenzo Di Filippo states that the robotic arm was mounted on one floating platform, while the target – described in the demonstration as a yellow object – was on another. First, the system was supposed to autonomously find the target and approach using sensors on the platform itself, and then the “tentacles” wrapped around the object and pulled it closer, mimicking the initial phase of capture in orbit.
Unlike SKYWALKER, where the main story was the transition of the algorithm from simulation to hardware, GRASP had to optimize the mechanics of touch as well: how much force it can apply, how to avoid pushing the target away, and how to measure the success of the capture. ESA mentions in the campaign report that testing was held in November 2025 and that the team, along with hardware challenges, had to manage interfaces and network communication in the laboratory. Despite this, the planned tests were performed, and the gripper results were rated as encouraging, although efficiency varied depending on the type of target, which is expected for technology in an early development stage.
Broader context: from “flat floor” to complex missions in orbit
Although ORBIT at first glance looks like a laboratory floor and a few “gliding” platforms, its role is to connect three levels of verification: theory and simulation, ground testing with realistic limitations, and finally application in space. In the field of orbital robotics, where precise maneuvers near other objects are performed, every millimeter and every millisecond of delay can be crucial.
Comparatively, interest in technologies like autonomous “crawling” over large structures is growing with plans for increasingly large systems in orbit: from service platforms to structures that would be assembled in space, where the dimension limitation of rocket launches is one of the main obstacles. In that sense, SKYWALKER’s goal – a robot that independently moves over a structure – is not just an academic exercise, but a potential building block for future assembly and maintenance systems.
On the other hand, GRASP’s adhesive gripper targets a problem that already burdens the space community today: the growing amount of debris and the need for reliable capture of non-cooperative objects. The ORL is, according to ESA’s descriptions, one of the laboratories where such mechanisms can be verified in realistic “free-floating” conditions before demonstrations in orbit are considered.
V-STARS, meanwhile, shows that a “microgravity analog” can also be used for humans: how perceptual strategies change when the body cannot rely on gravity as a constant frame of reference. In crewed missions, this is a question that enters the domain of work safety and operational efficiency, from moving through a module to performing precise tasks in the first days of flight.
“Creative ideas become reality”: what ESA highlights as the result
ORL lead Marti Vilella, in ESA’s review of the campaign, said that in the laboratory he sees students arriving with bold concepts and leaving after achieving an ambitious and unique result themselves. In his words, such projects are not just a technological step forward, but also a springboard for professional careers.
Program coordinator Laura Borella emphasized that the ESA Academy Experiments Programme follows teams from concept and design to testing and execution in advanced facilities related to gravity research, with constant mentorship from ESA experts. She particularly pointed out that the program is not closed to STEM only, but is open to students of various profiles – from design and psychology to communications and business – because interdisciplinarity often improves the usability, quality, and reach of projects.
For ESA, this is also a way to test fresh ideas in a real environment, but also to create a new generation of engineers and researchers who understand how space projects are developed, documented, and led through risks. And for students, ORBIT is that rare place where the difference between “it worked in simulation” and “it works on hardware” is seen immediately – and where, with mentorship, it can be turned into the next step of development.
Sources:- European Space Agency (ESA) – report on three student teams on ORBIT and description of experiments ( link )- ESA – technical description “Orbit Flatfloor” (dimensions, flatness tolerances, VICON system, and purposes) ( link )- ESA – Orbital Robotics Laboratory (ORL) page within the ESA Academy Experiments Programme (description of capabilities and microgravity simulations) ( link )- ESA – “About the ESA Academy Experiments programme” (program phases and platforms) ( link )- ESA – “SKYWALKER tests robotic crawling at ESTEC” (details of the campaign in late September 2025 and description of two algorithms) ( link )- ESA – “V-STARS pioneers neuroscience at ESA’s Orbital Robotics Lab” (approvals, campaign duration, and description of the VR task) ( link )- ESA – “Gripping the Future: Students Test Gecko-Inspired Robotic Arm…” (campaign in November 2025 and gripper result) ( link )
Find accommodation nearby
Creation time: 4 hours ago