Advancing Perception and Threat Identification of Autonomous Vehicles via Reconfigurable Phononic Metastructures

Project #3.28 begins in 2026.

Traditional methods such as radar have proven insufficient to identify and locate small drones flying low. Moreover, radar is expensive and requires significant power. Instead, acoustic methods (inexpensive and very low power) have increasingly been used in combat with excellent success. For instance, acoustic systems can identify drones from up to 5 km based on the sound from the drones’ motors. While promising, these systems use very directive parabolic dishes to capture sounds produced by far away sources, oriented along the direction of highest acoustic sensitivity, making them easily detectable targets. It is thus of critical importance to develop automated perception systems conformal with the autonomous vehicle while effectively scanning all directions from which a threat might arise. As an alternative to acoustic parabolic dishes, phononic crystals have been shown

The goal of this research is to explore a novel reconfigurable phononic structure idea that can be easily implemented with broad-range scanning and real-time Dirac cone tuning, advancing perception and identification of acoustic wave arriving from controllable directions (drone noise) to locate and classify threats at kilometer ranges.

To pursue such a goal, several research questions are identified: (a) How to create phononic structure lattices conformal to the vehicle that are easily reconfigurable to achieve Dirac cone manipulation in a broad range of directions? (b) How to engineer adaptive material lattices that are sensitive to sound arriving in air from controllable directions? (c) How to integrate the material lattices with automated algorithms to perceive and classify sound sources?

Task 1 - Create a novel reconfigurable, conformable metastructure with origami lattice and metamaterials, simple and inexpensive to realize and sensitive to sound arriving from tunable directions.

Task 2 - Develop novel digital twin and digital engineering (DE)-based machine-learning to co-design parameters to achieve desired Dirac cones.

Task 3 - Process the sound captured by the origami receiver using an automated classifier framework based on digital engineering (DE) principles.

3.28