Conformal Active Beamforming Metasurfaces

Abstract

A traditional sensor is designed to be sensitive and respond only to a single target input, necessitating the use of multiple sensors to sense multiple target inputs. From a traditional sensor viewpoint, a sensor sensitive to various stimuli simultaneously is considered useless. However, by changing this conception, we propose the development of a machine-learned multi-modal sensor, where a single sensor can extract multiple signals corresponding to various stimuli without the need for multiple sensors. This proposal encompasses an integrated bottom-up research approach covering materials, fabrications, devices, and algorithms for development to real-world applications. By elucidating the inherent characteristics of the sensor and their correlation with responses to multiple stimuli, we aim for the single multi-modal sensor to achieve a data-rich signal output. This approach addresses the issue of increased systemic complexity usually associated with enhanced functionality in conventional sensor systems. Additionally, it includes an analysis of sensor utility from a machine learning perspective, presenting key criteria for future sensor development. This work will fundamentally alter the concept of traditional sensor research.

Document Details

Document Type
DoD Grant Award
Publication Date
Feb 05, 2025
Source ID
FA23862414090

Entities

People

  • Kamran Ghorbani

Organizations

  • Air Force Office of Scientific Research
  • RMIT University
  • United States Air Force

Tags

Fields of Study

  • Engineering

Readers

  • Distributed Systems and Data Platform Development
  • Theoretical Analysis.

Technology Areas

  • AI & ML
  • AI & ML - Neural Networks