Loading presentation...

Present Remotely

Send the link below via email or IM


Present to your audience

Start remote presentation

  • Invited audience members will follow you as you navigate and present
  • People invited to a presentation do not need a Prezi account
  • This link expires 10 minutes after you close the presentation
  • A maximum of 30 users can follow your presentation
  • Learn more about this feature in our knowledge base article

Do you really want to delete this prezi?

Neither you, nor the coeditors you shared it with will be able to recover it again.


Make your likes visible on Facebook?

Connect your Facebook account to Prezi and let your likes appear on your timeline.
You can change this under Settings & Account at any time.

No, thanks

Robotics and Tensegrity--Jeff Friesen 10-30-13

No description

Tehseen Lazzouni

on 4 November 2013

Comments (0)

Please log in to add your comment.

Report abuse

Transcript of Robotics and Tensegrity--Jeff Friesen 10-30-13

Robotics, Tensegrity and Controls
Previous duct climbing designs
Why the Intelligent Robotics Group at NASA is Interested in Tensegrity
Design Highlights
Control of Duct Climbing Tetrahedral Tensegrity (DuCTT)
What is Tensegrity?
Jeffrey Friesen
Pay special attention in MAE 140 and 143A/B
Take MAE 143C
Go to office hours, the proffesors at UCSD are good people to know
Rigidly jointed
Very heavy, because they require large actuators to move limbs for climbing
If offered a research oppurtunity, say YES
My first research experience was in the SHM lab
Performed some basic FEA analysis
Conducted some experiments on propagation of ultrasonic surface waves through metal joints
Helped design and build SHEMA, a structure for SHM benchmarking

Try to find an interesting 199 project
It turns research into class credit
If you can publish results from your 199 project,even better
Publishing as an undergraduate looks awesome on grad school and fellowship applications
Getting involved in research at UCSD
None of this seems overly related to my current work but my experience in that research group has been extremely valuable to me.
Learning more about robotics and control
Formulating a Compelling Research Question Involving Tensegrity
Tensegrity structures first created as art
by Kenneth Snelson
Buckminster Fuller adopted them for their favorable structural characteristics
Biotensegrity: Tensegrity as the architechture of life
Tensegrity at the cellular level
"As I see it, this type of structure, at least in its purest form is not likely to prove highly efficient or utilitarian. As the engineer Mario Salvadori put it to me many years ago, "The moment you tell me that the compression members reside interior of the tension system, I can tell you I can build a better beam than you can."
Many molecular and cellular structures have been modeled as tensegrity structures, where the tensile members are often some form of intermolecular force.
Control is a non-trivial problem
Because the structures are inherently compliant and oscillatory traditional control techniques are often difficult to implement
These same characteristics make them ideal matches for bio-inspired control policies
These bio-inspired policies can be much more robust in unknown environments
If we want to build a robot with human-like capabilities we should start with a human-like structure
Tensegrity Structures
are commonly very lightweight and deployable
Some tensegrity projects that were already underway In the lab at Ames
SuperBall Robot
Tetraspine Robot
Brian Tietz designed a CPG controller inspired by cockroach brain to control a snakelike robot made of tetrahedral links
Mission Concept
Send robot to titan
Drop from orbit
Mobility structure also cushions payload during landing
Control Strategies
For more information on this topic check out Vytas Sunspiral's blog at: magicalrobot.org
Atıl İşçen developed a controller using evolutionary learning algorithms which requires minimal sensory information

Jérémie Despraz created a control policy based on central pattern generators which requires extensive sensory data but is very robust in unknown environmnets
Hardware Prototypes
Ken Caluwaerts built ReCTER at Ghent University
Jonathan Bruce, Andrew Sabelhaus, and Ken are currently working at Ames to build a much more powerful SuperBall Prototype, the first rod is pictured at the right
Idea: Use actuated tetrahedral frame for succesive wedging
Easy to build
Not really tensegrity
Can't turn
Very slow
Implementing a spine-like nested tetrahedral topology
Nested Tetrahedron Topology
Same successive wedging as single tetrahedron but with a compliant 6 DOF flexible joint to allow for complex maneuvering
Design Iterations
How it climbs
Controller performance in hardware
Inverse Kinematics of tensegrity structures
Assume you know how you want your structure to be positioned and what external forces are present
Necessary forces can be calculated using some well defined linear algebra and from there rest lengths of strings determined
Since DuCTT has 6 DOF and 8 actuated cables, the solution is non-unique
Minimizing force density via a quadratic optimization in the cables with a constraint that force density is still greater than 0 yields an energy efficient answer
This can be done very quickly in matlab or python (>100 Hz)
Testing Controller performance in simulation
Physics model constructed using NASA Tensegrity Robotics Toolkit(NTRT) built on top of bullet, a video game physics simulator
This will soon be open source software
Leap Motion module used to intuitively control the high DOF motion of the robot
Evaluating Controller Performance
In Hardware
In Simulation
Topology somewhat resembles shoulder anatomy
Minimal cross-sectional area may allow for a robustified version to move through duct systems while they are online
Non-Backdrivable Gearing on strings
Future Work
New mechanical design
Cleaner Electronics
Force sensing for impedance control
Tensegrity Linear Actuator
How to only use one motor
We created spiral cams to reel in cables at appropriate rates according to relative changes in length
How to build a linear actuator using tensegrity concepts
By countrwinding the central axis, the angle of the four bar linkage changes
Has a larger % change in length than off the shelf linear actuators
How it works
Is capable of motion and works as expected but high friction due to imprecise machining makes it's motion jerky and inconsistent
Full transcript