Dark Ship Detection in the Indian Ocean Region (IOR)
Dr Cdr Arnab Das [Director, Maritime Research Center Pune]
The Indian Ocean Region (IOR), is gaining strategic relevance in the 21st century with more and more extra-regional powers maintaining strategic presence in the region. The political instability in the IOR, is also ensuring security concerns and particularly the state and non-state actors being equally active and at times in cahoots, ensuring a deadly cocktail for maritime forces to counter. The state-of-the-art underwater technology available even with the subversive elements is a matter of serious concern and the security establishment needs to up the ante to always remain on top of the game.
The rising submarine proliferation in the IOR, is one dimension that demands serious consideration by our security establishment and cannot be countered by conventional means.
The rising submarine proliferation in the IOR, is one dimension that demands serious consideration by our security establishment and cannot be countered by conventional means. The area coverage required for the entire Exclusive Economic Zone (EEZ) and beyond is massive, to be able to monitor, presence of dark ships through the ongoing conventional methods. The tropical littoral waters of the IOR further add to the challenges as the sonars deployed for underwater surveillance present substantial sub-optimal performance due to the acoustic propagation fluctuations.
The dark ships are the general category of platforms including surface and sub-surface vessels that could potentially cause disruption to Good Order at Sea.
The dark ships are the general category of platforms including surface and sub-surface vessels that could potentially cause disruption to Good Order at Sea. The Automatic Identification System (AIS), is a system that provides details of the surface vessels and their voyage based on the regulations of the International Maritime Organization (IMO). The AIS data has been extensively used for generating Maritime Domain Awareness (MDA) to counter multiple security concerns. Any vessel not maintaining expected course and speed, as per certain predetermined norms are put under the suspected category and then segregated by the security forces for scrutiny. The conventional means for generating MDA can only detect surface ships with AIS data feed. However, it is grossly inadequate for sub-surface targets like Submarines and also other submersibles. The Underwater Domain Awareness (UDA) requires far different efforts to overcome the limitations of the ongoing MDA and also the tropical littoral challenges of sonar deployment.
The dark ship detection has been a matter of concern globally post the 9/11 incident in the US and more locally in the IOR, post the 26/11 incident in Mumbai.
The dark ship detection has been a matter of concern globally post the 9/11 incident in the US and more locally in the IOR, post the 26/11 incident in Mumbai. There have been multiple techniques proposed by various agencies globally to detect dark ships, however, none of them could comprehensively undertake the task. Some of these techniques are as follows:
AIS based high end data analytics for MDA. An Israel based company is the market leader for this technique, however it suffers from the limitation of being able to detect only surface platforms with AIS data inputs.
Synthetic Aperture Radar Image Analysis, combined with AIS data analysis supported by Machine Learning and Artificial Intelligence algorithms. A Finnish company is promoting this technique with lot of success. This again contributes to the MDA only and fails in effective UDA.
Visible Infrared Imaging Radiometer Suite (VIIRS) mounted on satellites for monitoring of Illegal, Unreported and Unregulated (IUU) fishing activities. These IUU fishing vessels often double up as vessels of opportunity for the non-state actors determined to mount large scale damage. VIIRS combined with AIS data analysis has also been attempted to enhance dark ship detection accuracy. This again is limited to IUU fishing only and fails in case of UDA.
Radio Frequency (RF) Analysis through a constellation of satellite monitoring varied RF emissions from the surface vessels. The company promoting this technology has been accused of privacy breach, and also only enhances the MDA capabilities and is limited for effective UDA.
Passive Acoustic Monitoring (PAM) uses acoustic hydrophones to record emissions from the vessels and data analytics along with signal processing to detect the dark ships. This although can undertake UDA but gets overwhelmed by the underwater signal distortions particularly in the tropical littoral waters of the IOR and also the high volume of shipping traffic.
The Maritime Research Centre (MRC), Pune has developed a very unique technique for dark ship detection that can comprehensively address the challenges of effective UDA in the IOR. The method has been validated with limited real data with encouraging results.
All these techniques are unable to address the specific requirement of the IOR and demand an alternate look at the way ahead. The Maritime Research Centre (MRC), Pune has developed a very unique technique for dark ship detection that can comprehensively address the challenges of effective UDA in the IOR. It is a two stage model that extensively uses the data analytics algorithms and passive acoustic sensing to identify the dark ships.
The first stage is the spatio-temporal low frequency mapping using the AIS data feed in the region combined with the site specific underwater channel model. The entire region is divided into grids of specific size, depending upon the desired spatial resolution and the shipping traffic data is extracted from the AIS data base for each grid. The radiated noise from each grid is computed and passed through the underwater channel model to obtain the low frequency ambient noise of the region. The low frequency ambient noise map, thus generated accounts for the background noise levels at the specific location. The ambient noise so generated gets updated every six minutes (the minimum update rate of the AIS input) thus creating the realistic spatio-temporal map. This also provides the Signal-to-Noise Ratio (SNR) in the region that can be used to deploy the hydrophones at the appropriate location.
In the second stage, the acoustic sensor is deployed at the location with minimal SNR to ensure effective range of detection. The recorded acoustic ambient noise is compared with the predicted spatio-temporal low frequency ambient noise map input to compute the anomaly. In case the vessel is not accounted for in the AIS data while computing the spatio-temporal map, it will appear as an anomaly. This will thus allow detection of all vessels not accounted for in the AIS database.
The detected dark ship input from the underwater acoustic sensor can further be analysed for classification and also localization. The sensor array configuration in the horizontal and vertical planes and signal processing has been appropriately configured to generate very high end analysis outputs. State-of-the-art, Machine Learning (ML) and Artificial Intelligence (AI) algorithms have been deployed for enhancing the computational efficiency and accuracy of the results. The method has been validated with limited real data with encouraging results.
Multiple aspects of this study were undertaken at MRC, Pune as part of a summer internship project by Mr. Pranjal Kapoor from BITS Pilani, during his six weeks attachment. He was ably supported by Mr. Shridhar Prabhuraman, research coordinator at MRC. The details of his findings are available at (Webpage | Research Note ) and there are two components to his deliverables. One is the research note that summarizes the state-of-the-art in the domain to establish his specific novel contribution and the second is the detailed report that provides his findings with results and analysis. There is scope to take forward this unique effort and translate into a tool for defence application, effective UDA framework realization, blue economic policy formulation and oceanographic studies in the IOR and beyond.