MIT scientists create cube robots that can change shape in space
If you face a choice between sending a swarm of full-sized individual robots into space or a large team of smaller robotic modules, you can choose the latter.
Modular robots, like those shown in films such as "City of Heroes", hold particular promise thanks to their abilities to self-assemble and reconfigure. But for all their ambitious aims of rapid, reliable deployment in areas spanning space exploration, search and rescue, and shape-changing, modular robots built today are still somewhat clumsy.
Typically, they are built from a set of large, expensive motors to enable movement, which calls for much-needed attention to more scalable architectures, both in number and size.
Scientists from the Massachusetts Institute of Technology's Computer Science and Artificial Intelligence Laboratory (MIT CSAIL) turned to electromagnetism, the magnetic fields created by the flow of electric current, to avoid the usual assembly of bulky, expensive actuators into separate units.
Instead, they embedded small, easy-to-manufacture and inexpensive electromagnets into the edges of the cubes, which repel and attract, allowing the robots to rotate and move around one another, as well as quickly change shape.
ElectroVoxel has a side length of about 60 mm, and its magnets consist of a ferrite core (a type of magnetic conductor that looks like a small black tube) wrapped in copper wire, with a total cost of just $0.60. Inside each cube are tiny circuit boards and electronics that pass current through the right electromagnet in the right direction.
Unlike traditional hinges, which require mechanically joining two elements, ElectroVoxel is fully wireless, which greatly simplifies maintenance and the production of large-scale systems.
To better visualize how a group of blocks will look while interacting, the scientists used a software planner that visualizes reconfigurations and calculates the underlying electromagnetic distributions. A user can manipulate up to a thousand cubes with just a few mouse clicks, or use predefined scripts that encode several sequential turns.
The system does allow the user to control the blocks' fate within reasonable limits: they can change speed, highlight magnets, and display the necessary movements to avoid collisions. You can instruct the blocks to take different shapes, such as a chair or a sofa.
Cheap small blocks are especially beneficial for microgravity conditions, where any structure you want to launch into orbit must fit inside the rocket used to launch it. After initial tests of ElectroVoxel, the scientists observed true weightlessness during flight tests in zero-gravity conditions, with the overall goal of improving space exploration tools, such as fuel-free reconfiguration or changing the inertial properties of a spacecraft.
For example, by using fuel-free actuation, there is no need to carry extra fuel for reconfiguration, which addresses many problems related to launch mass and volume.
Thus, there is hope that this reconfiguration method could help many future space endeavors: expanding and replacing space structures over multiple launches, building temporary structures to assist with inspecting spacecraft and aiding astronauts, and future iterations of the cubes acting as standalone sorters of storage containers.
“ElectroVoxel shows how to design a fully reconfigurable system and introduces our scientific community to the challenges that must be solved to have a fully functional modular robotic system in orbit. This research demonstrates how simple it is to create, operate and maintain rotating cubes with electromagnetic actuation, enabling a flexible, modular and reconfigurable system that may inspire the development of intelligent components for future research missions”
For the blocks to move, they must follow a specific sequence, like small uniform pieces in Tetris. In this case, the polarization sequence consists of three steps: launch, travel and catch, with each phase involving a traveling cube (which moves), a source (from which the traveling cube launches) and a destination (which catches the traveling cube).
Software users can specify which cube to rotate in which direction, and the algorithm will automatically calculate the sequence and the addresses of the electromagnetic targets needed to do so (repel, attract or switch off).
As for future work, moving from space back to Earth is the natural next step for ElectroVoxel, which will require more detailed modeling and optimization of these electromagnets for reconfiguration against gravity on our planet.
“When building a large, complex structure, you do not want to be limited by the availability and experience of the people assembling it, the size of your vehicle, or adverse environmental conditions at the assembly site. While these principles hold true on Earth, they are very important for building things in space. If you had structures that assemble from simple uniform modules, you could eliminate many of these problems. So, although the potential advantages in space are especially large, the paradox is that the favorable dynamics provided by microgravity mean some of these problems are actually easier to solve here, since in space even tiny forces can set large objects in motion. By applying this technology to solve real short-term problems in space, we hope to create technology for future use on Earth as well”
Nisser wrote the paper with Leon Cheng and Yashaswini Makaram of MIT CSAIL, University of Calgary computer science assistant professor Ryo Suzuki, and MIT professor Stefanie Mueller. They will present their work at the 2022 International Conference on Robotics and Automation. The work was partially supported by the MIT Space Exploration Initiative.
