Research Overview

Research opportunities:

The ARL conducts basic and applied research in shallow-water acoustics and underwater technology. We focus on a few areas of research: underwater acoustic sensors & signal processing, ambient noise imaging, autonomous platforms for underwater sensing, underwater communication networks, and bioacoustics. The research is heavily cross-linked with plenty of interaction across the research areas.

Selected research topics and technologies:

Acoustic sensors and signal processing

  • Signal processing in impulsive noise: Snapping shrimp produce a cacophony of impulsive broadband noise in warm coastal waters around South East Asia, Australia, Mediterranean and Central America. Signal processing designed for Gaussian noise is poorly suited for these environments. We develop robust signal detection, beamforming, communication, channel equalization and error correction techniques that work well in impulse noise. more...

  • Lightweight towed arrays: Traditional towed arrays towed by ships are large and heavy. With the increasing popularity of smaller platforms such as autonomous underwater vehicles (AUVs) and unmanned surface vehicles (USVs), the interest in lightweight towed arrays is rapidly growing. We have developed thin towed arrays from 8 mm to 20 mm in diameter, and successfully used them from AUVs and other platforms. more...

  • Fibre laser hydrophone arrays: Distributed feedback fibre laser sensor (DFB-FLS) hydrophone technology is an attractive candidate for building lightweight towed arrays. We have successfully developed pressure-compensated DFB-FLS hydrophone prototypes with good performance up to 5 kHz. more...

  • Deep sea exploration: In collaboration with the Keppel-NUS corporate laboratory, we plan to explore how to use AUVs for large-scale deep sea surveying. The goal is to fuse data from cameras, side-scan sonars, multibeam sonars, sub-bottom profilers, etc to rapidly survey a mineral deposit and estimate its volume using adaptive statistical sampling strategies. more...

Ambient noise imaging

  • ROMANIS: Just like we use ambient light to see things around us, why not use ambient noise in the ocean for sensing? ROMANIS is a broadband acoustic camera built at the ARL, and successfully used for ambient noise imaging (ANI), passive target detection/ranging, and mapping of the underwater environment. more...

Autonomous vehicles and platform technologies

  • PANDA: The PANDAs are autonomous underwater nodes that can be deployed and recovered easily without diver support. While we use them as acoustic recorders, communication network nodes, communication relays or underwater navigation beacons, one could potentially add environmental or other sensors on them quite easily. more...

  • STARFISH: The small team of autonomous robotic fish (STARFISH) project was conceptualized to explore the use of multiple cooperative AUVs for underwater sensing and monitoring applications. To support this exploration, the STARFISH AUVs were developed as low-cost, open-architecture AUVs with a high degree of modularity and reconfigurability. This makes them an excellent research platform for testing of novel algorithms and research payloads. more...

  • SwarmBot: Swarm intelligence is the collective behaviour of decentralised, self-organising systems that typically consist of a population of simple agents interacting locally with one another and with their environment. To explore these principles in small teams of underwater agents, we are developing a number of low-cost underwater robots called as SwarmBots. more...

  • Persistent marine and fresh-water monitoring: We believe cost-effective monitoring of water bodies is possible with simple, unattended, autonomous platforms that work collaboratively to sense their environment, provide realtime data, and allow interactive mission control. To achieve this vision, we are developing a fleet of minimally propelled robotic platforms for near-persistent environmental monitoring in littoral waters and in-land water bodies. more...

  • Automated recovery of AUVs: Arguably, the hardest part in automated launch and recovery of AUVs is the capture of the AUV for recovery. The AutoLARS project was conceptualized to capture an AUV underwater autonomously by having the capture hoop move and intercept the AUV. more...

Underwater communication and sensing networks

The Unet project project strives to develop technologies and promote community collaboration in the area of underwater networks:

  • Unet-2 modem: The Unet-2 Modem is a software-defined modem developed for underwater communication and networking research. It is designed for flexibility, allowing novel modulation schemes/algorithms, forward error correction codes, and network protocols to be tested at sea. more...

  • UnetStack: UnetStack is an embedded Java/Groovy network stack for underwater networks. It is based on the open-source fjge lightweight agent framework, and provides a basic set of agents that allow an underwater network to be deployed. Designed for extensibility, UnetStack allows additional agents for optimized protocols to be rapidly developed, tested and deployed. The stack can easily be integrated with most underwater modems. more...

fjge: The framework for Java and Groovy agents (fjge) was initially developed to support STARFISH's agent-based command and control system, and later also adopted in the Unet stack, SwarmBot, NUSwan and other projects. The lightweight Java framework is now available as open-source software for the community to use and develop further. more...

  • UnetSim: UnetSim is an underwater network simulator based on the UnetStack implementation. It allows Unet agents and protocols to be simulated in realistic channel conditions, with minimum effort. The simulator is designed to be easy to learn and use, and to allow agent implementations to be shared between deployment environment and simulation environment. Essentially, once a protocol is developed and tested in simulation, it is ready to be deployed and tested at sea in any UnetStack-compatible modem. more...

Bioacoustics

  • Dolphin biosonar: Dolphins are able to investigate their environment through echolocation in conditions where vision is limited. Cross-modal matching-to-sample studies have shown that object shape can be directly perceived through echolocation in a holistic manner, similar to the direct shape perception of shape through vision. No man-made sonar operating in a comparable frequency band can perform this feat. We are investigating the dolphin's ability to recognize shape through echolocation with an aim to understand the underlying signal processing that the animal employs. more...

  • Humpback whale vocalization: We investigated vocalizations of humpback whales that come to the wintering grounds around the Hawaiian islands each year. Humpback whales produce "songs" that typically last about 10-12 minutes. The function of humpback whale song remains unclear although several explanations have been offered over the years. We developed instrumentation that allows us to make field measurements on the source level, directionality and source location of the sounds produced by overlaying "acoustic contours" on video images of these animals. more...

  • Acoustic bandwidth compression: We developed a real-time acoustic bandwidth compression algorithm that enables acoustic signals with large bandwidth to be compressed into a bandwidth that humans can hear. The application of this algorithm is to enhance the diver experience, by enabling him or her to appreciate the rich soundscape the underwater world has to offer. We have demonstrated that a stereo implementation of the algorithm enables the diver to localize the sound source. This further enhances the experience and may potentially improve dive safety. more...
 

ARL