Avoiding singularities in robotics

From 33C3_Public_Wiki
Revision as of 02:15, 28 December 2016 by Petaflot (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search
Description Presenting a type of hybrid, serial-parallel bio-inspired robot geometry suitable for large and small robotic manipulators that do not exhibit the usual singularities. It has little or no limitations in speed or trajectory and the payload-to-weight ratio is far better than current serial robots, while the workspace is larger than most parallel robots known today.

If you have knowledge in 3D software like solidworks and like to animate a model, or - even better, if we find a way to hack a prototype during days 1,2,3 of the congress - get in touch with me!

Slides http://62.220.136.212/33c3_slides.pdf
Website(s) https://www.youtube.com/watch?v=zlGCurgsqg8, https://www.youtube.com/watch?v=0ZBD2tcKOU4
Tags singularity, robot, manipulator, anthropomorphism, bio-inspired
Person organizing Petaflot
Contact: the_robot@mon-net.org
Language en - English
en - English
Duration 5
Desired session Day 4
Desired timeframe begin, middle, end

refresh

Since the middle of the XXth century, the first programmable automata made their way into factories. Generic manipulator arms with usually six degrees of freedom and serial rotary joints became the norm. However, this architecture (and other) suffer from singularities, which must be properly dealt with in order to avoid the destruction of the robot, the tool, the part being machined or all of these. Tool speed near the singularities is limited while joints sustain considerable acceleration, so the effective workspace of an robot arm is often limited to a volume devoid of these singularities. In addition to this, serial manipulators are heavy in regard to the workload, slow and relatively inaccurate because of the serial configuration. Many unsuccessful attempts have been done at solving or displacing the singularity problem ; the most recent attempts to mimic human or animal limbs as closely as possible remain unusable because passive, elastic elements need to be added in series with the actuators and results in shaky systems. I will present a type of bio-inspired robot geometry suitable for large and small robotic manipulators that do not exhibit the singularities known today. They have little limitations in speed or trajectory, and the payload-to-weight ratio is far better than current serial robots.

If you have knowledge in 3D software like solidworks and like to animate a model, or - even better, if we find a way to hack a prototype during days 1,2,3 of the congress - get in touch with me!