CHANNELS

Aerospace

Automotive

Automated & Autonomous Vehicles

Battery & Electrification Technology

Defense

Drone TV

Electronics

Manufacturing & Prototyping

Materials

Medical

Motion Control

RF & Microwave Technology

Robotics & Automation

Sensors & IoT

Test & Measurement

Aerospace
Automotive
Automated & Autonomous Vehicles
Battery & Electrification Technology
Defense
Drone TV
Electronics
Manufacturing & Prototyping
Materials
Medical
Motion Control
RF & Microwave Technology
Robotics & Automation
Sensors & IoT
Test & Measurement
Aerospace & Defense
Search Results

Deformable 'Super Ball Bot' Could Be NASA's Next Titan-Exploring Robot

Researchers Vytas SunSpiral and Adrian Agogino from NASA's Ames Research Center are developing a tensegrity-based lander and planetary exploration robot. Tensegrity structures consist of rigid components such as hollow, cylindrical rods connected by flexible materials like elastic cable. The robot design, called the Super Ball Bot, might be able to land on an extraterrestrial surface without assistance - absorbing most of the shock of impact itself and saving the mass needed for more complex landing gear. Since Mars has a thin atmosphere that would likely require a ball bot to carry a parachute, SunSpiral and Agogino are considering Saturn's moon Titan as the robot's first target. Titan has an atmosphere that can slow a spacecraft's descent enough so that the robot can be dropped from 100 kilometers or more above the surface.

Topics:
Aerospace Automation Motion Control Robotics

More From SAE Media Group

    Transcript

    00:00:00 we're investigating how to use these really novel tensile uh soft compliant sort of more biologically inspired robots to explore space by shortening and changing the lengths of the cables the whole thing can roll around the planetary surface [Music] today we're going to be looking at what we call our super ball bot the

    00:00:26 tensegrity robot that we're designing that is able to both land on a planetary surface absorbing impact and is also able to move around and explore the planet this shows the unpacking from a very compact position you know to the full robot which is very important in space travel because um your payload fairing volume is at previous premiums so if you can pack this tightly

    00:00:51 you can pack multiple of them or you can make a very cheap mission out of it in traditional nasa mission landing is one of the most difficult expensive and one of the most unreliable things about a mission this is kind of a fundamentally simple landing system if you can survive a hard landing and you keep that system you can survive almost anything you can go off little cliffs

    00:01:13 you can go down you know steep train and so forth so it really gives you a very secure robust system what you see here in the landing is the central payload and it's that little sphere in the middle of all the rods it's protected from impact forces of landing by the elastic absorption of energy in this tensegrity

    00:01:35 structure much like we use our muscles to move our bodies around you we are going to be shortening and lengthening the cables of these structures to create motion by changing the dynamic balance of tensions in the system a benefit and a curse of these robots is there's a whole lot of control points and a lot of flexibility so that's really great in that they can go up hills they can

    00:01:59 handle bumps they can handle an even train but it's also very difficult to control so instead of the traditional control kind of top down we know how to control something we just tell it what to do instead our primary approach has been to evolve a control the best we can hope for is to give it lots of options of what it may do so we can select hundreds

    00:02:19 if not thousands of different options then some are good and some are bad at the beginning most of them are bad but you slowly take the good ones like evolution you replicate the good ones make small changes then eventually the good ones get better and better and you know out of all the thousands of bed controllers you actually evolve a few good ones one of the interesting

    00:02:39 questions is how does a structure like this move through a field of rubble and rocks and whatever you might encounter on the surface of the another planet and so this was an initial first uh pass at saying what does it look like for a robot to this type of robot to encounter a bunch of obstacles and move through them

    00:02:58 one of the advantages of a tensegrity robot like we're designing is it's a very compliant and forgiving system and so we're trying to maximize the ability of the system to move through environments reactively and then it makes your high-level navigation control problem a lot easier so you can imagine actually putting four or five all in one aeroshell and all

    00:03:22 unpacking you know very nice and neat and so you could have a mission where you can have four or five or possibly if you made them small dozens or even hundreds all going at the same time and then that clip i believe showed high-level algorithms of coordination so you can imagine many of these you know that they're all robust in themselves and then they can all coordinate with

    00:03:41 each other and perform science quickly and also reliably you know if a few of them don't make it it's okay the others will coordinate and make up for that when you're exploring another planet what drives the mission is the scientific instruments that you want to bring to that planet to explore to ask basic questions we're really looking at this from a fundamental principles

    00:03:59 perspective how do you how do you design and control these types of devices for now we're building prototypes that are about a meter in diameter or each rod is a meter long because that's a that's a nice natural size that you can get lots of components for motors and and you can fit in your controlled electronics electronics without too much effort we're currently

    00:04:20 designing with that target in mind is that how would you build a system like this that could deliver a 75 kilogram payload to the surface of titan but in terms of its potential you could be building very large versions of these or very small versions the intuition is that as it scales bigger and bigger the mass of the system will scale approximately linearly the same

    00:04:40 properties hold for for a very small version of this but uh you're obviously your payload your scientific instruments that you're going to bring with you are going to be more limited based on size and mass availability our whole approach has started from the ground up as being a messy system the integrity itself is messy the controls are messy it's

    00:05:01 oscillatory the evolution's messy we're dealing with lots of evolved components but part of the advantage of that because at every step of the way we've had an expectation that things are messy things are difficult when we throw little obstacles in its way it's really not a big deal we've already encountered so much which isn't true in a lot of research a

    00:05:21 lot of theory where you have this ideal world you've assumed perfect vacuums or perfect frictionless environments or perfect precision and everything and then the world world doesn't work but we've never assumed any of these mainly because we could never assume any of these it was from the ground up you know very noisy and messy so that actually gives us an advantage

    00:05:44 in the future we can really do all these things and we can expect it will probably work [Music] you