CDS&E: A Computational Framework for Parsimonious Sonar Sensing

点击翻译按钮获取中文摘要

Autonomous sensing systems play an increasingly important role in health care, manufacturing, transportation, disaster response, national security and other activities. Sensing systems in platforms such as autonomous vehicles often require clear lines-of-sight and are inefficient. They generate large amounts of data that cannot be processed by small platforms with limited onboard computation. This project lays the foundation for a new transformative data-centered efficient sensing platform based on bats' bio-sonar. The researchers will develop a computational framework that facilitates estimation of environmental parameters for sensing and navigation--enabling enhanced performance in complex environments and with reduced hardware requirements. This enables the use of autonomous devices for a broad scope of applications ranging from cooperative robots to medical devices and micro-air vehicles. The project will also offer multidisciplinary training to university students and will engage K-12 students in outreach activities.This project will develop a computational approach to advance the discovery and understanding of parsimonious sonar sensing in complex environments. The researchers will perform large-scale simulations to generate densely vegetated environments and vegetation echoes; develop novel statistical models to extract echo signatures and learn their relation to environmental parameters and design dynamic algorithms for sensing and navigation. This computational approach provides enabling tools--validated using simulated and experimental data--to advance the community's ability to develop adaptive models for various sensing scenarios and enable novel sensing paradigms. Simulation algorithms and algorithms for sensing and navigation will be integrated into open source packages and made available on GitHub. The conceptual framework will be published in appropriate journals.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.