Learning modular policies for robotics

Neumann, Gerhard; Daniel, Christian; Paraschos, Alexandros; Kupcsik, Andras; Peters, Jan

doi:10.3389/fncom.2014.00062

Cited by 17 publications

(12 citation statements)

References 29 publications

(61 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Along the same direction, the Probabilistic Movement Primitive (ProMP) approach in [4], [5], uses a model-free approach to encode a distribution over trajectories and analytically derive a stochastic feedback controller to reproduce the given trajectory distribution. This allows for flexibility over the possible motion generation such as spatial and temporal rescaling, combination and blending of the modeled motion primitives (MPs).…”

Section: Related Workmentioning

confidence: 99%

Learning assistive teleoperation behaviors from demonstration

Havoutis

Calinon

2016

2016 IEEE International Symposium on Safety, Security, and Rescue Robotics (SSRR)

View full text Add to dashboard Cite

Abstract-Emergency response in hostile environments often involves remotely operated vehicles (ROVs) that are teleoperated as interaction with the environment is typically required. Many ROV tasks are common to such scenarios and are often recurrent. We show how a probabilistic approach can be used to learn a task behavior model from data. Such a model can then be used to assist an operator performing the same task in future missions. We show how this approach can capture behaviors (constraints) that are present in the training data, and how this model can be combined with the operator's input online. We present an illustrative planar example and elaborate with a non-Destructive testing (NDT) scanning task on a teleoperation mock-up using a two-armed Baxter robot. We demonstrate how our approach can learn from examples task specific behaviors and automatically control the overall system, combining the operator's input and the learned model online, in an assistive teleoperation manner. This can potentially reduce the time and effort required to perform teleoperation tasks that are commonplace to ROV missions in the context of security, maintenance and rescue robotics.

show abstract

Section: Related Workmentioning

confidence: 99%

Learning assistive teleoperation behaviors from demonstration

Havoutis

Calinon

2016

2016 IEEE International Symposium on Safety, Security, and Rescue Robotics (SSRR)

View full text Add to dashboard Cite

show abstract

“…Each cell in this architecture can also be modeled as a complicated neural system and more neural systems are coupled in this network. DMPs, as a mathematical dynamics modifier, work with actor-critic RL mechanism (Kober and Peters, 2010 ) to optimize/learn a locomotor system. In the future work, a sophisticated memory system, which includes short-term memory and long-term memory, is required in our system to map the contextual factors into parameter space.…”

Section: Discussionmentioning

confidence: 99%

A novel approach to locomotion learning: Actor-Critic architecture using central pattern generators and dynamic motor primitives

Lowe

Ziemke

2014

Front. Neurorobot.

View full text Add to dashboard Cite

In this article, we propose an architecture of a bio-inspired controller that addresses the problem of learning different locomotion gaits for different robot morphologies. The modeling objective is split into two: baseline motion modeling and dynamics adaptation. Baseline motion modeling aims to achieve fundamental functions of a certain type of locomotion and dynamics adaptation provides a “reshaping” function for adapting the baseline motion to desired motion. Based on this assumption, a three-layer architecture is developed using central pattern generators (CPGs, a bio-inspired locomotor center for the baseline motion) and dynamic motor primitives (DMPs, a model with universal “reshaping” functions). In this article, we use this architecture with the actor-critic algorithms for finding a good “reshaping” function. In order to demonstrate the learning power of the actor-critic based architecture, we tested it on two experiments: (1) learning to crawl on a humanoid and, (2) learning to gallop on a puppy robot. Two types of actor-critic algorithms (policy search and policy gradient) are compared in order to evaluate the advantages and disadvantages of different actor-critic based learning algorithms for different morphologies. Finally, based on the analysis of the experimental results, a generic view/architecture for locomotion learning is discussed in the conclusion.

show abstract

“…In this paper, we unify and complement our prior work [33,36,37] on ProMPs. Note that the reference [33] contains only a brief summary of our work on ProMPs presented in the context of an overview paper that spans over multiple topics. Therefore, [33] provides less information than the corresponding conference papers.…”

Section: Introductionmentioning

confidence: 96%

“…Note that the reference [33] contains only a brief summary of our work on ProMPs presented in the context of an overview paper that spans over multiple topics. Therefore, [33] provides less information than the corresponding conference papers. In this paper, we present much more details which are necessary to reproduce the results.…”

Section: Introductionmentioning

confidence: 99%

Using probabilistic movement primitives in robotics

et al. 2017

Self Cite

View full text Add to dashboard Cite

Movement Primitives are a well-established\ud paradigm for modular movement representation and\ud generation. They provide a data-driven representation\ud of movements and support generalization to novel situations,\ud temporal modulation, sequencing of primitives\ud and controllers for executing the primitive on physical\ud systems. However, while many MP frameworks exhibit\ud some of these properties, there is a need for a uni-\ud fied framework that implements all of them in a principled\ud way. In this paper, we show that this goal can be\ud achieved by using a probabilistic representation. Our\ud approach models trajectory distributions learned from\ud stochastic movements. Probabilistic operations, such as\ud conditioning can be used to achieve generalization to\ud novel situations or to combine and blend movements in\ud a principled way. We derive a stochastic feedback controller\ud that reproduces the encoded variability of the\ud movement and the coupling of the degrees of freedom\ud of the robot. We evaluate and compare our approach\ud on several simulated and real robot scenarios

show abstract

Learning modular policies for robotics

Cited by 17 publications

References 29 publications

Learning assistive teleoperation behaviors from demonstration

Learning assistive teleoperation behaviors from demonstration

A novel approach to locomotion learning: Actor-Critic architecture using central pattern generators and dynamic motor primitives

Using probabilistic movement primitives in robotics

Contact Info

Product

Resources

About