The Effects of Learning in Morphologically Evolving Robot Systems

@article{Luo2021TheEO,
  title={The Effects of Learning in Morphologically Evolving Robot Systems},
  author={Jie Luo and Jakub M. Tomczak and Agoston E. Eiben},
  journal={Frontiers in Robotics and AI},
  year={2021},
  volume={9}
}
Simultaneously evolving morphologies (bodies) and controllers (brains) of robots can cause a mismatch between the inherited body and brain in the offspring. To mitigate this problem, the addition of an infant learning period has been proposed relatively long ago by the so-called Triangle of Life approach. However, an empirical assessment is still lacking to-date. In this paper, we investigate the effects of such a learning mechanism from different perspectives. Using extensive simulations we… 

Evolving modular soft robots without explicit inter-module communication using local self-attention

VSRs, aggregations of mechanically identical elastic blocks, are focused on, where the same neural controller is used inside each voxel, but without any inter-voxel communication, hence enabling ideal conditions for modularity: modules are all equal and interchangeable.

References

SHOWING 1-10 OF 50 REFERENCES

Gait-learning with morphologically evolving robots generated by L-system

A system where modular robots can create offspring that inherit the bodies of parents by recombination and mutation is set up, and it is shown that the evolution plus learning method does not only lead to a higher fitness level, but also to more morphologically evolving robots.

Lamarckian Evolution of Simulated Modular Robots

The possibility of bootstrapping infant robot learning through employing Lamarckian inheritance of parental controllers is investigated, and it is observed that changing the way controllers are evolved also impacts the evolved morphologies.

Evolving-Controllers Versus Learning-Controllers for Morphologically Evolvable Robots

This work investigates an evolutionary robot system where (simulated) modular robots can reproduce and create robot children that inherit the parents’ morphologies by crossover and mutation and shows that the learning approach does not only lead to different fitness levels, but also to different (bigger) robots.

Comparing lifetime learning methods for morphologically evolving robots

Bayesian Optimization and Differential Evolution are applied as learning algorithms and compared on a test suite of different robot bodies and they are found to work on all possible robot morphologies and be efficient.

Scalable co-optimization of morphology and control in embodied machines

A technique for ‘morphological innovation protection’ is demonstrated, which temporarily reduces selection pressure on recently morphologically changed individuals, thus enabling evolution some time to ‘readapt’ to the new morphology with subsequent control policy mutations.

Morpho-evolution with learning using a controller archive as an inheritance mechanism

This work proposes a framework that combines an evolutionary algorithm to generate body-plans and a learning algorithm to optimise the parameters of a neural controller where the topology of this controller is created once the body-plan of each offspring body- plan is generated.

Evolution and Learning in an Intrinsically Motivated Reinforcement Learning Robot

The most important results show that systems using both evolution and learning outperform systems using either one of the two, and that systems evolving internal reinforcers for learning building-block skills have a higher evolvability than those directly evolving the related behaviors.

Memetic robot control evolution and adaption to reality

It is demonstrated that Lamarckian evolution is effective in improving the performance of robot controller evolution, and that the same learning process on the physical robot efficiently reduces the negative impact of the simulation-reality gap.

Embodied intelligence via learning and evolution

Deep Evolutionary Reinforcement Learning (DERL) is introduced: a novel computational framework which can evolve diverse agent morphologies to learn challenging locomotion and manipulation tasks in complex environments using only low level egocentric sensory information.

Real-world evolution adapts robot morphology and control to hardware limitations

This paper applies real world multi-objective evolutionary optimization to optimize both control and morphology of a four-legged mammal-inspired robot, and shows that evolution under the different hardware limitations results in comparable performance for low and moderate speeds.