Building a Universal Theory of Multi-Sensory Integration

Abstract

Efficiently locating the source of a chemical plume is an unresolved engineering challenge, with significant implications. For example, source-localization is a prerequisite for neutralizing chemical or biological weapons. Flying insects are remarkably adept at following odor plumes and may provide inspiration for novel algorithms. One reason underlying insects’ performance is their utilization of myriad sensory modalities. Historically it has been difficult to determine how insects integrate olfactory and other information, but new neurogenetic tools (optogenetics) and the nearly complete connectome of the adult Drosophila brain now make this possible. The objective of this proposal is to experimentally investigate how insects integrate wind and olfactory information together, over time, and with vision, to efficiently follow turbulent odor plumes. To achieve this objective, I have developed a novel approach which uses optogenetics to create a virtual odor experience for flying flies, giving me unprecedented control of their olfactory experience in a manner that is completely independent of the wind. My behavioral experiments will provide new insight into how three different sensory modalities are integrated together, over time. By doing this work in a genetic and neural model system, my results will pave the way towards understanding how the brain controls this process. Meanwhile, my simulations will explore the impact of flies’ behavioral decisions at large spatial scales, allowing me to build a general theory of sensory integration that can be compared across a diversity of plume tracking species. My insights will shed new light on how brains make decisions in complex environments and will help to inspire novel algorithms for myriad robotic applications

Document Details

Document Type
DoD Grant Award
Publication Date
Jan 21, 2022
Source ID
FA95502110122XX0

Entities

People

  • Floris van Breugel

Organizations

  • Air Force Office of Scientific Research
  • United States Air Force
  • University of Nevada, Reno

Tags

Readers

  • Agent-Based Social Robotics and Mobile-Assisted Learning in Virtual Environments.
  • Neuroscience
  • Theoretical Analysis.

Technology Areas

  • AI & ML
  • AI & ML - Autonomous Systems
  • Autonomy
  • Autonomy - Autonomous System Control
  • Biotechnology