Semester of Graduation

Spring 2026

Degree

Master of Science in Mechanical Engineering (MSME)

Department

Department of Industrial and Mechanical Engineering

Document Type

Thesis

Abstract

Autonomous Underwater Vehicles (AUVs) are untethered robotic platforms used for tasks such as seafloor mapping, infrastructure inspection, and environmental monitoring. Recent technological advances have produced smaller, more affordable platforms, broadening access to research teams and small companies alike. This miniaturization comes at the cost of them handling drawbacks associated with a more compact machine such as reduced battery capacity as well as limited processing and sensing capabilities. These constraints make small-sized marine vehicle’s reliability critical as they can cause malfunctions, making the loss of a vehicle more likely. Actuator faults are particularly consequential as unintended and unstable control in an already dynamic underwater environment can easily ruin data collection or lead the platform into more disastrous situations. Recent advances in the development of Artificial Neural Networks (ANNs) have led to several algorithms to detect actuator faults under noisy conditions. However, these neglect to address the need for extremely lightweight solutions capable of monitoring actuator health real-time. This study explores using lightweight bio-inspired Spiking Neural Networks (SNNs) to detect thruster faults in AUVs to address this gap. In this thesis, simulated vehicle state health information with artificial actuator fault injections is used to train several feedforward and deep convolutional SNN architectures. Various signal to spike encoding schemes are considered such as the temporal Step Forward Encoding (SFe) method and the novel Banded Ensemble Encoding (BEe) technique. Experimental results show that there is a need to supply multiple signals to SNNs if SFe inputs are used. Using the BEe technique causes more effective function approximation in feedforward SNNs and efficient feature extraction capabilities in deep convolutional SNN. For the deep convolutional SNNs implemented, low module activity but high fault detection capabilities are observed.

Date

3-20-2026

Committee Chair

Barbalata

LSU Acknowledgement

1

LSU Accessibility Acknowledgment

1

Download

Share

COinS