Recent technological advances have opened new possibilities for the development of robotic systems, including spacecraft for the exploration of other planets. These new systems could ultimately contribute to our understanding of our galaxy and the unique characteristics of the many celestial objects it contains.
Researchers at RoMeLa, the Robotics and Mechanisms Laboratory at University of California Los Angeles (UCLA) have designed SPLITTER (Space and Planetary Limbed Intelligent Tether Technology Exploration Robot), a multi-robot system for planetary exploration comprised of two sub-10kg quadrupedal robots connected by a tether.
Their proposed system, set to be presented at the IEEE Aerospace Conference (AeroConf) 2025, could jump on the surface of the moon or asteroids, while also collecting data from its surroundings. The work is also published on the arXiv preprint server.
“The inspiration for this study stemmed from the challenges of locomotion and attitude control (3D orientation control) in low-gravity environments such as the moon or asteroids,” Yusuke Tanaka, first author of the paper, told Tech Xplore.
“Traditional rovers are large, heavy, and slow, limiting the area and efficiency of planetary exploration.
“Aerial solutions like drones can observe larger surface areas than rovers, but they are impractical due to the lack of atmosphere on the moon or asteroids. Our work explores an alternativeβleveraging dynamic, successive jumping gaits combined with inertial morphing for in-flight stability.”
The research team at RoMeLa previously introduced a limbed climbing robot, which they dubbed SCALER (Spine-enhanced Climbing Autonomous Legged Exploration Robot), which was designed for planetary exploration.
When they tested this robot, however, they found that it moved slowly, both when walking and climbing surfaces. They thus set out to explore methods to enhance its efficiency, without drastically changing its overall structure.
“The primary objective of this paper was to demonstrate attitude control through inertial morphing by adjusting inertia with changes in limb configurations and tether length, making the SPLITTER robot a mass-efficient and scalable solution for planetary exploration,” said Tanaka.
“Ours is one of the first (if not the first) approaches to achieve attitude control through inertial morphing with a Model Predictive Controller (MPC).”
In their new study, Tanaka and his colleagues introduced an inertial morphing mechanism based on an MPC. This mechanism regulates the orientation of their robot while it is flying and improves its stability.
“The inertial morphing mechanism leverages the Tennis Racket Theorem (Dzhanibekov effect), which describes how objects with asymmetric inertia can experience spontaneous flips when rotating around their intermediate axis,” explained Tanaka.
“Our inertial morphing-based attitude controller makes use of this principle to enable aggressive stabilization of the mid-air flight stability in a controlled manner.”
The researchers’ paper outlines the design for the new robotic system, dubbed SPLITTER, including actuator specifications required for its operation. The robot essentially consists of two small four-legged robots, called Hemi-SPLITTERs, that are connected by a tether to form a dumbbell-like structure.
“The robot’s limbs are powerful enough to jump under reduced gravity, yet they require stable mid-flight control,” said Tanaka. “Instead of relying on heavy reaction wheels or gas thrust for attitude control, SPLITTER dynamically changes its inertia by adjusting the length of the tether and the positioning of its limbs.”
The new system designed by this research team could have various advantages over many previously proposed robots for planetary exploration. One of its key advantages is its mass efficiency, which eliminates the need for dedicated attitude control hardware such as gas thrusters, reaction wheels, or wings, and enhanced agility.
“The tether mechanism is useful for planetary exploration as well,” said Tanaka. “For example, one Hemi-SPLITTER can go into a crater or cave while the other side is anchored to support. Our successful jumping locomotion is effective since it can store jumping energy in rotation, meaning it can keep accelerating with each jump.”
The SPLITTER robot’s design and the mechanism underpinning its locomotion could be particularly well-suited for the exploration of low-gravity environments. In these environments, conventional wheeled robots have been found to be inefficient, while aerial robots are not always easy to deploy.
“Our findings demonstrate that SPLITTER can control the mid-air robot’s posture through inertial morphing technique with MPC in our simulation,” said Tanaka. “We employed MPC to show that the system can regulate its angular velocity/orientation and maintain stability without requiring external forces or momentum wheels.”
The researchers envision the deployment of their robotic systems as a robot swarm, which could efficiently traverse and explore a vast and unstructured environment. The MPC-based inertial morphic mechanism they developed could potentially also be applied to other robots, as well as satellites and spacecraft, to improve their stability in space.
“Our future research will focus on experiments of SPLITTER in high-fidelity simulation to further validate inertial morphing MPC, in which more accurate physics simulation and robot motion analysis are possible,” added Tanaka. “This paper is a part of Project SPLITTER, our foundational work that researches necessary technologies to develop SPLITTER hardware.”
Currently, the RoMeLa team at UCLA is working on further enhancing their robot’s hardware. For instance, they are focusing on the development of new actuators and sensing mechanisms for SPLITTER, which could broaden its capabilities.
“Our lab’s expertise lies in legged robots, mostly for ground locomotion on earth. From what we have learned from our decades of research developing these systems, we are now applying them to space applications, which is again introducing a whole new set of difficult challenges.”
Dr. Dennis Hong, the PI for the project and director of RoMeLa told Tech Xplore, “With new challenges, we create new solutions and new knowledge which make this job more meaningful and interesting.”
Β© 2025 Science X Network
