Noise and the Reality Gap: The Use of Simulation in Evolutionary Robotics

@inproceedings{Jakobi1995NoiseAT,
  title={Noise and the Reality Gap: The Use of Simulation in Evolutionary Robotics},
  author={Nick Jakobi and Phil Husbands and Inman Harvey},
  booktitle={ECAL},
  year={1995}
}
The pitfalls of naive robot simulations have been recognised for areas such as evolutionary robotics. [] Key Method This included detailed mathematical models of the robot-environment interaction dynamics with empirically determined parameters. Artificial evolution was used to develop recurrent dynamical network controllers for the simulated robot, for obstacle-avoidance and light-seeking tasks, using different levels of noise in the simulation.
A Neural Network-based kinematic and light-perception simulator for simple robotic evolution
TLDR
ANNs were employed to simulate the motion dynamics of a mobile robot steered using differential steering, as well as the interaction of two light sensors onboard the robot with a light source in its vicinity, indicating that ANNs show definite promise as robot simulators.
Filling the reality gap: Using obstacles to promote robust gaits in evolutionary robotics
TLDR
This paper investigates if the addition of a set of small obstacles in the simulated environment can help promote more robust gaits when transferred to a real world robot.
The transferability of evolved hexapod locomotion controllers from simulation to real hardware
TLDR
This work used Evolutionary Robotics to evolve simple open-loop locomotion controllers for a real-world hexapod (six-legged) robot in a physics engine-based simulation and indicated that the transferability of these evolved controllers was greatly aided by the simulator optimization, noise injection and incorporation of torque limits in the simulator.
Evolutionary Robotics and the Radical Envelope-of-Noise Hypothesis
TLDR
New ways of thinking about and building simulations upon which fitness assessments of evolving controllers can be performed are outlined and a potential methodology for building simulations in which evolving controllers are forced to satisfy these conditions if they are to be reliably fit is developed.
On the use of simulation in robotics: Opportunities, challenges, and suggestions for moving forward
TLDR
It is posited that well-validated computer simulation can provide a virtual proving ground that in many cases is instrumental in understanding safely, faster, at lower costs, and more thoroughly how the robots of the future should be designed and controlled for safe operation and improved performance.
On Mimicking the Effects of the Reality Gap with Simulation-Only Experiments
TLDR
It is argued it is unnecessary to assume reality is more complex than simulation models for the effects of the reality gap to occur, and it is shown that performance drop and rank inversion can occur if one automatically designs control software in simulation using a model and then assesses it in simulation on another model—what the authors call a pseudo-reality.
Simulation-only experiments to mimic the effects of the reality gap in the automatic design of robot swarms
TLDR
Whether, to observe the effects of the reality gap, it is necessary to assume that the control software is designed in a context that is simpler than the one in which it is evaluated is investigated.
Evolutionary Robotics Applied to Hexapod Locomotion: a Comparative Study of Simulation Techniques
TLDR
It is indicated that ANN-based simulators offer a superior alternative to widely-used physics simulators in ER for the locomotion task considered, and are vastly more computationally efficient than the physics-based simulator.
Behaviour Trees for Evolutionary Robotics: Reducing the Reality Gap
TLDR
This paper shows the first application of the Behaviour Tree framework to a real robotic platform using the Evolutionary Robotics methodology to improve the intelligibility of the emergent robotic behaviour as compared to the traditional Neural Network formulation.
Blurred Vision: Simulation-Reality Transfer of a Visually Guided Robot
TLDR
The success in evolving in simulation a robot controller incorporating only distal visual environment input data and displaying the same behaviours in both simulation and the real world, goes some way to addressing the argument that evolution is suitable only for toy problems.
...
...

References

SHOWING 1-10 OF 27 REFERENCES
Selection for Wandering Behavior in a Small Robot
TLDR
It is observed that evolution was an effective means to program the robot's behavior and progress was characterized by sharply stepped periods of improvement, separated by periods of stasis that corresponded to levels of behavioral/computational complexity.
Artificial Life and Real Robots
TLDR
A new abstraction for behavior-based robot programming which is specially tailored to be used with genetic programming techniques is introduced, which will be necessary to automatically evolve programs that are one to two orders of magnitude more complex than those previously reported in any domain.
Evolution Versus Design: Controlling Autonomous Robots
  • P. HusbandsI. Harvey
  • Computer Science
    Proceedings of the Third Annual Conference of AI, Simulation, and Planning in High Autonomy Systems 'Integrating Perception, Planning and Action'.
  • 1992
This paper sets out and justifies a methodology for the development of the control systems, or ‘cognitive architectures)) of autonomous mobile robots. It will be argued that the design b y hand of
How to Evolve Autonomous Robots: Different Approaches in Evolutionary Robotics
TLDR
Different examples of applications of evolutionary robotics to real robots are shown by describing three different approaches to develop neural controllers for mobile robots by compared with each other.
Evolving Electronic Robot Controller that Exploit Hardware Resources
TLDR
This paper reasons that constraints can be removed when using artificial evolution, releasing huge potential even from small circuits, including the first reported evolution of a real hardware control system for a real robot.
Integrating reactive, sequential, and learning behavior using dynamical neural networks
This paper explores the use of dynamical neural networks to control autonomous agents in tasks requiring reactive, sequential, and learning behavior. We use a genetic algorithm to evolve networks
Seeing the light: artificial evolution, real vision
TLDR
Results from a specialised piece of visuo-robotic equipment which allows the evolution of control systems for visually guided autonomous agents acting in the real world are described, showing some of these control systems to exhibit a surprising degree of adaptiveness when tested against generalised versions of the task for which they were evolved.
Intelligence without Representation
Intelligence Without Reason
TLDR
It is claimed that the state of computer architecture has been a strong influence on models of thought in Artificial Intelligence over the last thirty years.
Introduction to Control Theory
Introduction PART I: ORDINARY CONTROL THEORY 1. Ordinary linear systems 2. Feedback control and integral action 3. Stability of feedback systems 4. Root locus diagrams 5. Transfer functions for
...
...