Natural swarms and crowds: observation and modeling
Abstract
Natural swarms and crowds: observation and modeling Project Summary ONR Technical Point of Contact: Christiane N. Duarte O ce of Naval Research, Code 333WP One Liberty Center, Room 663 875 N. Randolph Street, Suite 1425 Arlington, VA 22203-1995 703.696.6947 christiane.duarte@navy.mil Recently, there has been increased interest in deriving bio-inspiration from collective animal behavior for engineering applications. In particular, many researchers have used the biological backdrop when deriving theoretical control models for autonomous vehicle groups [5, 6, 12, 11]. However, there is a lack of high-resolution data from which insights into collective behavior can be drawn and models can be validated [9, 1]. The present state-of-the-art method for high resolution experimental measurement of animal groups is two dimensionsal (2D) tracking [8], multi-camera stereoscopic imaging [13, 3] or gps tracking [14]. Imaging-based methods are attractive as they are non-invasive to the natural phenomena under investigation [13, 3, 8]. However, accurate three dimensional (3D) position and sensory data for large swarms of more than 30 individuals is needed for improved modeling that can both capture natural behavior and be applicable to autonomous vehicle groups [11]. The lack of such data has limited current models to a combination of three main types: (1) metric range (interaction with all neighbors within a xed distance) [4], (2) topological neighbors ( xed number of nearest neighbors [2]), and (3) Voronoi range (geometric tessellation of nearest neighbors [7]). These models lack information about sensory inputs, such as sight, and there is growing evidence that sensory-based models are more useful and accurate [11]. In fact, it has been known for some time that the wave-like response times of mocking birds in large groups requires information about individuals far away [10]. I propose a research paradigm shift through novel observation methods and sensor-based modeling.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Aug 12, 2016
- Source ID
- N000141512687
Entities
People
- Tadd T Truscott
Organizations
- Office of Naval Research
- United States Navy
- Utah State University