Limits...
Walknet, a bio-inspired controller for hexapod walking.

Schilling M, Hoinville T, Schmitz J, Cruse H - Biol Cybern (2013)

Bottom Line: It has been tested by kinematic and dynamic simulations as well as on a number of six-legged robots.Walknet shows how a number of properties observed in insects may emerge from a decentralized architecture.Finally, we briefly address the ability of this decentralized reactive controller to form the basis for the simulation of higher-level cognitive faculties exceeding the capabilities of insects.

View Article: PubMed Central - PubMed

Affiliation: Department of Biological Cybernetics and Theoretical Biology, Bielefeld University, P.O. Box 100131, 33501 , Bielefeld, Germany. malteschilling@googlemail.com

ABSTRACT
Walknet comprises an artificial neural network that allows for the simulation of a considerable amount of behavioral data obtained from walking and standing stick insects. It has been tested by kinematic and dynamic simulations as well as on a number of six-legged robots. Over the years, various different expansions of this network have been provided leading to different versions of Walknet. This review summarizes the most important biological findings described by Walknet and how they can be simulated. Walknet shows how a number of properties observed in insects may emerge from a decentralized architecture. Examples are the continuum of so-called "gaits," coordination of up to 18 leg joints during stance when walking forward or backward over uneven surfaces and negotiation of curves, dealing with leg loss, as well as being able following motion trajectories without explicit precalculation. The different Walknet versions are compared to other approaches describing insect-inspired hexapod walking. Finally, we briefly address the ability of this decentralized reactive controller to form the basis for the simulation of higher-level cognitive faculties exceeding the capabilities of insects.

Show MeSH

Related in: MedlinePlus

A network diagram describing a leg controller that summarizes a number of behavioral observations as detailed in the text. The right hand side depicts the sensory input relevant for stance control, the left hand side correspondingly the sensory input required for the control of the swing movement. Pairs of units depicted in blue at the center (Protractor-Retractor, Levator-Depressor, Flexor-Extensor) which are coupled via mutual inhibition (T-shaped connections), represent an abstract version of the network shown by Schumm and Cruse (2006, their Fig. 7). For definition of joint angles , and  see Fig. 1. Two motivation units (swing, stance, marked in red) control the sensory input to the control network via inhibitory connections
© Copyright Policy - OpenAccess
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3755227&req=5

Fig5: A network diagram describing a leg controller that summarizes a number of behavioral observations as detailed in the text. The right hand side depicts the sensory input relevant for stance control, the left hand side correspondingly the sensory input required for the control of the swing movement. Pairs of units depicted in blue at the center (Protractor-Retractor, Levator-Depressor, Flexor-Extensor) which are coupled via mutual inhibition (T-shaped connections), represent an abstract version of the network shown by Schumm and Cruse (2006, their Fig. 7). For definition of joint angles , and see Fig. 1. Two motivation units (swing, stance, marked in red) control the sensory input to the control network via inhibitory connections

Mentions: The general architecture of Walknet. Only two leg controllers are shown (for details see Fig. 5). The upper part contains the motivation units (all marked in red) forming a heterarchical network influencing the procedures (black boxes, e.g., Swing-net, Stance-net, Target-net_fw, and Target-net_bw, representing the end point of the swing movement for forward walking and backward walking, respectively). Furthermore, there are “higher-level” motivation units as leg1, walk, as well as forward (fw) and backward (bw). A motivation unit able to control a coordination influence between leg1 and leg 2 is marked by r1. Motivation units form a recurrent neural network coupled uni- or bidirectionally by positive (arrowheads) and negative (T-shaped connections) influences. The lower part of the figure (dashed box muscles/body/environment) schematically depicts the “loop through the world”


Walknet, a bio-inspired controller for hexapod walking.

Schilling M, Hoinville T, Schmitz J, Cruse H - Biol Cybern (2013)

A network diagram describing a leg controller that summarizes a number of behavioral observations as detailed in the text. The right hand side depicts the sensory input relevant for stance control, the left hand side correspondingly the sensory input required for the control of the swing movement. Pairs of units depicted in blue at the center (Protractor-Retractor, Levator-Depressor, Flexor-Extensor) which are coupled via mutual inhibition (T-shaped connections), represent an abstract version of the network shown by Schumm and Cruse (2006, their Fig. 7). For definition of joint angles , and  see Fig. 1. Two motivation units (swing, stance, marked in red) control the sensory input to the control network via inhibitory connections
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3755227&req=5

Fig5: A network diagram describing a leg controller that summarizes a number of behavioral observations as detailed in the text. The right hand side depicts the sensory input relevant for stance control, the left hand side correspondingly the sensory input required for the control of the swing movement. Pairs of units depicted in blue at the center (Protractor-Retractor, Levator-Depressor, Flexor-Extensor) which are coupled via mutual inhibition (T-shaped connections), represent an abstract version of the network shown by Schumm and Cruse (2006, their Fig. 7). For definition of joint angles , and see Fig. 1. Two motivation units (swing, stance, marked in red) control the sensory input to the control network via inhibitory connections
Mentions: The general architecture of Walknet. Only two leg controllers are shown (for details see Fig. 5). The upper part contains the motivation units (all marked in red) forming a heterarchical network influencing the procedures (black boxes, e.g., Swing-net, Stance-net, Target-net_fw, and Target-net_bw, representing the end point of the swing movement for forward walking and backward walking, respectively). Furthermore, there are “higher-level” motivation units as leg1, walk, as well as forward (fw) and backward (bw). A motivation unit able to control a coordination influence between leg1 and leg 2 is marked by r1. Motivation units form a recurrent neural network coupled uni- or bidirectionally by positive (arrowheads) and negative (T-shaped connections) influences. The lower part of the figure (dashed box muscles/body/environment) schematically depicts the “loop through the world”

Bottom Line: It has been tested by kinematic and dynamic simulations as well as on a number of six-legged robots.Walknet shows how a number of properties observed in insects may emerge from a decentralized architecture.Finally, we briefly address the ability of this decentralized reactive controller to form the basis for the simulation of higher-level cognitive faculties exceeding the capabilities of insects.

View Article: PubMed Central - PubMed

Affiliation: Department of Biological Cybernetics and Theoretical Biology, Bielefeld University, P.O. Box 100131, 33501 , Bielefeld, Germany. malteschilling@googlemail.com

ABSTRACT
Walknet comprises an artificial neural network that allows for the simulation of a considerable amount of behavioral data obtained from walking and standing stick insects. It has been tested by kinematic and dynamic simulations as well as on a number of six-legged robots. Over the years, various different expansions of this network have been provided leading to different versions of Walknet. This review summarizes the most important biological findings described by Walknet and how they can be simulated. Walknet shows how a number of properties observed in insects may emerge from a decentralized architecture. Examples are the continuum of so-called "gaits," coordination of up to 18 leg joints during stance when walking forward or backward over uneven surfaces and negotiation of curves, dealing with leg loss, as well as being able following motion trajectories without explicit precalculation. The different Walknet versions are compared to other approaches describing insect-inspired hexapod walking. Finally, we briefly address the ability of this decentralized reactive controller to form the basis for the simulation of higher-level cognitive faculties exceeding the capabilities of insects.

Show MeSH
Related in: MedlinePlus