Acoustic Signature Data and ISCMMS Data Collection
Ebor Computing developed the Acoustic Signature Data and
ISCMMS* Data Collection (ASDIC) software for DSTO to produce
large self-noise databases used in automated, real time detection
of noise-faults in submarines and surface vessels.
This project involved:
- Collaborative research with DSTO on spectral comparison,
spectral clustering and automated noise-fault detection
- Technical report writing
- Signal processing
- C++ programming
- Java programming
- Real-time programming
- Client/server programming
- Developing algorithms for parallel processing
ASDIC continuously acquires platform status and navigation
sensor (ISCMMS) data and time-series data from the on-board
sensors. The time-series data are then optionally logged to
disk. Spectral data are created from the time-series data,
by performing FFT's using one thread per sensor, and the spectral
data can be logged to disk. The ISCMMS data can also be logged
to disk. The current ISCMMS state can then be used to retrieve
all spectral data for similar ISCMMS states from the baseline
for comparison with the current spectral data to determine
if a noise-fault has occurred. The results are then displayed
in a GUI. In future, there will also be onshore support for
adding logged data to the baseline database.
* Integrated Ship Control, Management and Monitoring System.
David O'Mara, Graham Gear, and Darryl R. McMahon, "Automated
self-noise monitoring and acoustic fault detection",
In Proceedings of Undersea
Defence Technology (UDT), October 8-10, Cheju, Korea,
ASDIC system overview.