A Combination of Machine Learning and Cerebellar-like Neural Networks for the Motor Control and Motor Learning of the Fable Modular Robot

Ismael Baira Ojeda; Silvia Tolu; Moises Pacheco; David Johan Christensen; Henrik Hautop  Lund

A Combination of Machine Learning and Cerebellar-like Neural Networks for the Motor Control and Motor Learning of the Fable Modular Robot

Ismael Baira Ojeda, Silvia Tolu, Moises Pacheco, David Johan Christensen, Henrik Hautop Lund

Copenhagen Center for Health Technology

Research output: Contribution to journal › Journal article › Research › peer-review

1156 Downloads (Pure)

Abstract

We scaled up a bio-inspired control architecture for the motor control and motor learning of a real modular robot. In our approach, the Locally Weighted Projection Regression algorithm (LWPR) and a cerebellar microcircuit coexist, in the form of a Unit Learning Machine. The LWPR algorithm optimizes the input space and learns the
internal model of a single robot module to command the robot to follow a desired trajectory with its end-effector. The cerebellar-like microcircuit refines the LWPR output delivering corrective commands. We contrasted distinct cerebellar-like circuits including analytical models and spiking models implemented on the SpiNNaker platform, showing promising performance and robustness results.

Original language	English
Journal	Journal of Robotics, Networking and Artificial Life
Volume	4
Issue number	1
Pages (from-to)	62–66
ISSN	2405-9021
Publication status	Published - 2017

Keywords

Motor control
Cerebellum
Machine learning
Modular robot
Internal model
Adaptive behavior

Access to Document

Jrnal 4 1 62 66Final published version, 1.01 MB

OpenUrl availability

Full text

HBP SGA1: Human Brain Project. Subproject 10 Neurorobotics Platform - SGA1
Tolu, S. (Project Coordinator), Lund, H. H. (PI), Capolei, M. C. (Other), Corchado Miralles, C. (Other) & Baira Ojeda, I. (Other)
01/04/2016 → 01/04/2018
Project: Research

Cite this

@article{e56f5586de4d4cffa0cfd5400eaecfa0,

title = "A Combination of Machine Learning and Cerebellar-like Neural Networks for the Motor Control and Motor Learning of the Fable Modular Robot",

abstract = "We scaled up a bio-inspired control architecture for the motor control and motor learning of a real modular robot. In our approach, the Locally Weighted Projection Regression algorithm (LWPR) and a cerebellar microcircuit coexist, in the form of a Unit Learning Machine. The LWPR algorithm optimizes the input space and learns theinternal model of a single robot module to command the robot to follow a desired trajectory with its end-effector. The cerebellar-like microcircuit refines the LWPR output delivering corrective commands. We contrasted distinct cerebellar-like circuits including analytical models and spiking models implemented on the SpiNNaker platform, showing promising performance and robustness results.",

keywords = "Motor control, Cerebellum, Machine learning, Modular robot, Internal model, Adaptive behavior",

author = "{Baira Ojeda}, Ismael and Silvia Tolu and Moises Pacheco and Christensen, {David Johan} and Lund, {Henrik Hautop}",

year = "2017",

language = "English",

volume = "4",

pages = "62–66",

journal = "Journal of Robotics, Networking and Artificial Life",

issn = "2405-9021",

publisher = "Atlantis Press",

number = "1",

}

A Combination of Machine Learning and Cerebellar-like Neural Networks for the Motor Control and Motor Learning of the Fable Modular Robot. / Baira Ojeda, Ismael; Tolu, Silvia; Pacheco, Moises et al.
In: Journal of Robotics, Networking and Artificial Life, Vol. 4, No. 1, 2017, p. 62–66.

Research output: Contribution to journal › Journal article › Research › peer-review

TY - JOUR

T1 - A Combination of Machine Learning and Cerebellar-like Neural Networks for the Motor Control and Motor Learning of the Fable Modular Robot

AU - Baira Ojeda, Ismael

AU - Tolu, Silvia

AU - Pacheco, Moises

AU - Christensen, David Johan

AU - Lund, Henrik Hautop

PY - 2017

Y1 - 2017

N2 - We scaled up a bio-inspired control architecture for the motor control and motor learning of a real modular robot. In our approach, the Locally Weighted Projection Regression algorithm (LWPR) and a cerebellar microcircuit coexist, in the form of a Unit Learning Machine. The LWPR algorithm optimizes the input space and learns theinternal model of a single robot module to command the robot to follow a desired trajectory with its end-effector. The cerebellar-like microcircuit refines the LWPR output delivering corrective commands. We contrasted distinct cerebellar-like circuits including analytical models and spiking models implemented on the SpiNNaker platform, showing promising performance and robustness results.

AB - We scaled up a bio-inspired control architecture for the motor control and motor learning of a real modular robot. In our approach, the Locally Weighted Projection Regression algorithm (LWPR) and a cerebellar microcircuit coexist, in the form of a Unit Learning Machine. The LWPR algorithm optimizes the input space and learns theinternal model of a single robot module to command the robot to follow a desired trajectory with its end-effector. The cerebellar-like microcircuit refines the LWPR output delivering corrective commands. We contrasted distinct cerebellar-like circuits including analytical models and spiking models implemented on the SpiNNaker platform, showing promising performance and robustness results.

KW - Motor control

KW - Cerebellum

KW - Machine learning

KW - Modular robot

KW - Internal model

KW - Adaptive behavior

M3 - Journal article

SN - 2405-9021

VL - 4

SP - 62

EP - 66

JO - Journal of Robotics, Networking and Artificial Life

JF - Journal of Robotics, Networking and Artificial Life

IS - 1

ER -

A Combination of Machine Learning and Cerebellar-like Neural Networks for the Motor Control and Motor Learning of the Fable Modular Robot

Abstract

Keywords

Access to Document

OpenUrl availability

Fingerprint

Projects

HBP SGA1: Human Brain Project. Subproject 10 Neurorobotics Platform - SGA1

Cite this