18-04-2011, 10:07 AM
Presented by:
Gayathri Mohan
[attachment=12265]
Introduction
A Robot is a mechanical or virtual ,artificial agent.
By its appearance or movements, conveys a sense that it has intent or agency of its own.
One of the long-standing challenges is achieving robust performance under uncertainty.
In this situation there is need for robots those who can heal their damages by themselves.
Self Healing Robots(SHR)
• Researchers of shr
Josh Bongard
Viktor Zykov
Hod Lipson
ERROR RECOVERY
Recovery from damage is a major concern in robotics.
A four legged robot automatically synthesizes a predictive model of its own topology.
These findings may help develop more robust robotics
What is self healing?
SELF MODELLING BRIEFLY
Phases in self healing
Self Model synthesis
Exploratory Action Synthesis
Target Behavior synthesis
• Algorithm used
Estimation-Exploration algorithm
Includes two phases
Estimation phase
Exploration phase
• estimation-exploration algorithm
This algorithm is essentially a co-evolutionary process comprising two populations.
One population is of candidate models of the target system.
The other population is of candidate unlabelled sentences.
Algorithm has two functions:
Damage hypothesis evolution (the estimation phase)
Controller evolution (the exploration phase).
• Flow chart of estimation-exploration phases
Exploration Phase
Controller Evolution- The exploration EA is used to evolve a controller for the simulated robot.
When the exploration EA terminates, the best controller from the run is transferred to and used by the physical robot.
Estimation Phase
Damage Hypothesis Evolution. Used to evolve a hypothesis about the actual failure incurred by the physical robot.
When the estimation EA terminates, the most fit damage hypothesis is supplied to the exploration EA.
The robot simulator is updated to model this damage hypothesis.
THE ROBOTS
Two hypothetical robots tested in this preliminary work
1)A quadrupedal robot
2)A hexapedal robot
The quadrupedal robot has eight mechanical degrees of freedom.
The hexapedal robot has 18 mechanical degrees of freedom:
• The Controllers
The robots are controlled by a neural network.
Neuron values and synaptic weights are scaled to lie in the range [−1.00, 1.00].
There is one output neuron for each of the motors actuating the robot
RESULTS OF ESTIMATION-EXPLORATION ALGORITHM
• The robots suffered a different damage scenario: the 10 scenarios
• Damage scenarios 1, 2, 3, 5 and 6 single gene in the genomes of the estimation EA.
• Scenarios 4, 7 and 8 represent compound failures, and require more than one gene to represent them.
• Case 9 represents the situation when the physical robot signals that it has incurred some damage, when in fact no damage has occurred.
• Case 10 represents an unanticipated failure: hidden neuron failure cannot be described by the estimation EA genomes.
ESTIMATION-EXPLORATION ALGORITHM OVERVIEW
1.Characterization of the target system
2. Initialization
3. Estimation Phase
4. Exploration Phase
5. Termination
6. Validation
Conclusion
The possibility of autonomous self-modeling has been suggested.
The processes demonstrate both topological and parametric self-modeling.
The ability to actively generate and test hypotheses
The robot's abilities suggest a similarity to human thinking as the robot tries out various actions to figure out the shape of its world.
These findings may help develop more robust robotics