Jan Wessnitzer

University of Edinburgh

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Trends

Author Order

Document Type

Co-Authors

Barbara Webb

University of Edinburgh

Julien Colomb

Humboldt-Universität zu Berlin

Jedrzej Blaszkiewicz

Freie Universität Berlin

Lutz Reiter

Freie Universität Berlin

Bjoern Brembs

Freie Universität Berlin

Joanna Young

University of Edinburgh

Chris Melhuish

University of the West of England

Michael Mangan

University of Sheffield

J. Douglas Armstrong

University of Edinburgh

Björn Brembs

University of Regensburg

Cooperative Institutions

University of Edinburgh

Freie Universität Berlin

Google (United States)

University of Missouri

University of Melbourne

University of the West of England

Bristol Robotics Laboratory

École Polytechnique Fédérale de Lausanne

École Normale Supérieure - PSL

ETH Zurich

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Field

Building Adaptive Structure Formations with Decentralised Control and Coordination

The MIT Press eBooks (2002)

Jan Wessnitzer Chris Melhuish

We investigate collective decentralised decision-making in simple mobile agents engaged in the task of structure formation where the radius of the formed structure is much larger than the communication radius of an individual agent. We are searching for minimal and local rules allowing adaptive behaviour in collectives of mobile agents. In previous experiments, we presented rules that govern the evolution of the inter-agent communication network toward some task-specific topologies that allow a decentralised collective of mobile agents to dynamically organise into simple geometric structures. In this paper, we study the problem of coordinating and controlling the global behaviour of the group of agents from their local interactions. The agents determine the successful termination of a structure formation and switch to another formation. In simulation we demonstrate global state changes and global state persistance in a decentralised collective of mobile agents.

10.7551/mitpress/3121.003.0063

Cite

Citations (0)

Multimodal sensory integration in insects—towards insect brain control architectures

Bioinspiration & Biomimetics (2006)

Jan Wessnitzer Barbara Webb

Although a variety of basic insect behaviours have inspired successful robot implementations, more complex capabilities in these 'simple' animals are often overlooked. By reviewing the general architecture of their nervous systems, we gain insight into how they are able to integrate behaviours, perform pattern recognition, context-dependent learning, and combine many sensory inputs in tasks such as navigation. We review in particular what is known about two specific 'higher' areas in the insect brain, the mushroom bodies and the central complex, and how they are involved in controlling an insect's behaviour. While much of the functional interpretation of this information is still speculative, it nevertheless suggests some promising new approaches to obtaining adaptive behaviour in robots.

Mushroom bodies

10.1088/1748-3182/1/3/001

Cite

Citations (133)

A model of associative learning in the mushroom body

Biological Cybernetics (2008)

Darren A. Smith Jan Wessnitzer Barbara Webb

Mushroom bodies

Associative learning

Facilitation

Neuropil

Stimulus (psychology)

Associative property

Neural facilitation

10.1007/s00422-008-0241-1

Cite

Citations (54)

Low Level Approaches to Cognitive Control

Cognitive systems monographs (2008)

Barbara Webb Jan Wessnitzer H. Rosano Matthew Szenher Markos Zampoglou

Artifact (error)

Associative property

Mushroom bodies

10.1007/978-3-540-88464-4_3

Cite

Citations (1)

Trajectory parameters of real flies and computer-generated data (grouping codes given below the graph).

Julien Colomb Lutz Reiter Jedrzej Blaszkiewicz Jan Wessnitzer Bjoern Brembs

Both A. Meander (turning angle divided by speed) and B. stripe deviation are similar in fly and computer-generated data. Red line denotes 45°, the mean value for computer-generated data. C–D. Centrophobism score for sitting (C) or for moving (D) is positive only for fly data. E. The distance traveled is different between the three types of data. Bars represent means and error bars standard errors, asterisks denote significant differences after a MANOVA analysis, n = 20 in each group.

Line (geometry)

10.1371/journal.pone.0042247.g005

Cite

Citations (0)

A Model of Non-elemental Associative Learning in the Mushroom Body Neuropil of the Insect Brain

Lecture notes in computer science (2007)

Jan Wessnitzer Barbara Webb Darren A. Smith

Mushroom bodies

Neuropil

Associative learning

Associative property

10.1007/978-3-540-71618-1_54

Cite

Citations (267)

Resonant neurons and bushcricket behaviour

Journal of Comparative Physiology A (2006)

Barbara Webb Jan Wessnitzer Sarah L. Bush Johannes Schul Jonas Buchli

Dynamics

10.1007/s00359-006-0199-1

Cite

Citations (29)

Calculation of angles and number of walks

Julien Colomb Lutz Reiter Jedrzej Blaszkiewicz Jan Wessnitzer Bjoern Brembs

A. The inner circle represents the platform, while the outer circle represents the arena and the light source (to scale). The bars represent the stripes (wide or narrow). Considering the movement from P₀ to P₁, α₀ is the absolute movement angle (similarly α₋₁} is the absolute movement angle of the movement P₋₁} to P₀). The turning angle γ can be calculated as α₀ - α₋₁}, it represents the change in direction at time 0. β is the “stripe deviation” angle, the angle from the movement to a vector going straight toward the middle of the stripe that is in the direction of the movement. In the “ltraj” object, α is assigned to P₀, β to P₁. Gray areas denote the sectors used to start and end a walk between stripes: a walk is counted for each passage from one gray area to the other. B. Trajectory example, zoomed on the platform size. The disposition of the stripes are at 90 and −90° as in A. Dots represent the position of the fly during the three first minutes of a test with narrow stripes, after down sampling to 10 Hz.

Position (finance)

Visual angle

10.1371/journal.pone.0042247.g003

Cite

Citations (0)

Activity calculations using different computations give similar results for endogenous locomotion and computer-generated data (grouping codes given below the graph)

Julien Colomb Lutz Reiter Jedrzej Blaszkiewicz Jan Wessnitzer Bjoern Brembs

Pause-activity patterns were determined using either a speed- (left, labeled 1) or a time threshold (right, labeled 2). A. Total activity time represents the time the animal is considered active. A2 was set to be similar in the computer-generated data. B. Duration of activity periods. Inset represents the same calculation as in B2 but considers only activity bouts leading to a displacement of 1 cm or more. C. Duration of the pause periods. D. Number of pauses. Bars represent means and error bars standard errors, n = 20 in each group.

Activity monitor

10.1371/journal.pone.0042247.g004

Cite

Citations (0)

A model of non-elemental olfactory learning in Drosophila

Journal of Computational Neuroscience (2011)

Jan Wessnitzer Joanna Young J. Douglas Armstrong Barbara Webb

Mushroom bodies

Antennal lobe

Associative learning

Associative property

10.1007/s10827-011-0348-6

Cite

Citations (59)