The iCub robot learns archery

Sep 29, 2010 by Lin Edwards report
Humanoid robot iCub learns the skill of archery. The learning algorithm, called ARCHER (Augmented Reward Chained Regression) algorithm, was developed and optimized specifically for problems like the archery training. Photo: Dr. Petar Kormushev

(PhysOrg.com) -- The humanoid robot iCub has learned a new skill: archery. After being taught how to hold a bow and shoot an arrow, it learned for itself how to improve its aim, and was so successful it could hit a bullseye after only eight trials.

The algorithm used to teach iCub is called the Augmented Reward Chained Regression (or ARCHER, naturally), and was developed especially for problems with specific knowledge about the goal to be achieved and which have a "smooth solution space." In archery, the greatest reward comes from hitting a bullseye, the center of the target.

This video is not supported by your browser at this time.
Humanoid robot iCub learns the skill of archery.

The iCub is an open-source about the same size as a child three and a half years old, or around 104 cm tall. It has 53 degrees of freedom (hence it is called the 53-DOF iCub) distributed about the head, torso, arms, legs, and hands. The distance between the target and the iCub was 3.5 meters.

The robot's hands were set manually to allow it to grip the bow, and one joint in the index finger was used to release the string and shoot the arrow. The algorithm controlled the posture of the left arm and the orientation of the right arm (string side).

ARCHER is a chained vector regression algorithm that uses experience gained from each trial to fine-tune the next attempt by modulating and coordinating the movements of the robot's hands. Movements of the arms are controlled by an inverse kinematics controller. After each shot a camera takes a picture of the target and an system based on Gaussian Mixture Models determines where the tip of the arrow hit the target by filtering the colored pixels of the picture based on their likelihood of belonging to the target or the arrow head. This information is then used as feedback for the algorithm.

Photo credits: Dr. Petar Kormushev

The algorithm was developed by Dr. Petar Kormushev and colleagues of the Italian Institute of Technology (IIT). The results of the research will be presented in December this year at the Humanoids 2010 conference in Nashville TN in the US.

Explore further: Brain-training for baseball robot

More information: Research paper: kormushev.com/papers/Kormushev_Humanoids-2010.pdf

Related Stories

iCub, the Toddler Robot (w/ Videos, Pictures)

Sep 09, 2009

(PhysOrg.com) -- A little humanoid robot called iCub is learning how to think for itself, bringing the world of science fiction to reality. The major goal of the "RobotCub" project is to study how humans learn ...

Robot teaches stroke survivors

Mar 15, 2010

Shaking hands with a robotic arm could be a new way to help stroke patients learn to use their arms again. Researchers writing in BioMed Central's open access Journal of NeuroEngineering and Rehabilitation re ...

Toy Robot to Solve Sudoku (w/ Video)

Sep 03, 2009

(PhysOrg.com) -- A Swedish programmer, Hans Andersson, has used a Lego Mindstorms NXT kit to develop a robot to solve Sudoku puzzles.

New Algorithm Ranks Sports Teams like Google's PageRank

Dec 15, 2009

(PhysOrg.com) -- Sports fans may be interested in a new system that ranks NFL and college football teams in a simple, straightforward way, similar to how Google PageRank ranks webpages. The new sports algorithm, ...

Recommended for you

Sony film mess reflects gadgets-entertainment gap

2 hours ago

Sony's iconic gadgetry and the star appeal of Hollywood may have appeared to be a perfect match when the electronics giant bought Columbia Pictures in 1989. A quarter century later, it's apparent that Sony ...

Streaming release of 'Interview' test for industry

Dec 25, 2014

Sony's "The Interview" has been a hacking target, a punchline and a political lightning rod. Now, with its release online at the same time it debuts in theaters, it has a new role: a test for a new kind of ...

User comments : 8

Adjust slider to filter visible comments by rank

Display comments: newest first

teledyn
3 / 5 (4) Sep 29, 2010
I thought that in archery the greatest reward comes from a mastery of the zen state of satori.
Lyte
1 / 5 (1) Sep 29, 2010
Let's teach it to use guns next...
Musashi
5 / 5 (2) Sep 29, 2010
I'd be more impressed if the robot autonomously reached for the string and pulled it back. Still neat.
kasen
4 / 5 (1) Sep 29, 2010
Wouldn't it have been easier, mechanically speaking, to use a small air cannon? I mean, unless the body and algorithm can be used for any hand-eye skill without extensive pre-programming, then the 53 DoF aren't really justified, for the purpose of proving learning ability.
plasticpower
not rated yet Sep 29, 2010
Next it will shoot all humans.
PPihkala
not rated yet Sep 29, 2010
It's more rewarding to use hard to control weapon like bow and arrow. The path of the arrow to target is not straight line and it's reach is also modulated by the strings position before release.
Vaughn
5 / 5 (1) Sep 30, 2010
There should be a law! Or three....
gurloc
5 / 5 (1) Oct 04, 2010
Oh good, first they were teaching robots to use deceit, http://www.physor...89.html, and now they are teaching them to use weapons.

This will end well...:-)

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.