Analysis of LRIT and AIS-via-Satellite trials data collected in January 2010
Executive Summary
In July 2009 and January 2010 the Australian Maritime Safety Authority (AMSA)
conducted trials of two emerging vessel tracking systems. The first of these, the
Automatic Identification System (AIS) has for a long time been a fundamental vessel
tracking system in the maritime domain. While AIS is conventionally a line-of-sight
radio broadcast system for communication between ships and between ships and shore
stations, recent technology developments have seen satellites adapted for reception of
AIS messages from Low Earth Orbit (LEO). The ‘AIS-via-Satellite’ (SAT-AIS) trial was
conducted during the same period as a trial of the Long Range Identification and
Tracking (LRIT) system. Like AIS, LRIT has been made a carriage requirement for
SOLAS class vessels. But, where AIS uses a line-of-sight radio broadcast, LRIT
communicates via satellite communications systems and disseminates information
through an international data exchange (IDE) outlined by the International Maritime
Organisation (IMO). It can therefore be said that LRIT and SAT-AIS utilise
fundamentally different technology systems to achieve their purpose.
In 2010, AMSA requested DSTO to provide an analysis of the SAT-AIS and LRIT data
collected during the January 2010 trials. Having briefly conducted its own analysis,
AMSA had identified that there were many instances observed where ships were seen
to report in one data set but not the other. In response, DSTO conducted both
quantitative and qualitative analysis, measuring both the size of the discrepancies seen,
and percentages of the combined vessel population observed by each system as well as
identifying a number of legitimate, systems, and data set factors that may help to
explain the differences between the reporting systems.
Two primary methods of computation were used. The first, called Type I, analysed the
pattern of reporting for vessels that were seen to report through both systems while
Type II statistics reported on the number of vessels seen through each system over the
course of a number of time periods. In addition to this form of analysis, a number of
qualitative analyses were conducted that have identified areas that may explain
discrepancies between the systems. This form of analysis included the examination of
reporting observations against vessel flag-state of registration, vessel-type, as well as
reporting on metrics from a number of different regions.
Results from the analysis produced the following comparative observations:
- Over the entire population of vessels AIS-via-Satellite observed a greater number of unique vessels (83% to 92%) than LRIT (31% to 40%). 19% to 28% of all vessels were seen by both systems. This suggests that AIS-via-Satellite has greater overall visibility of vessel traffic.
- For those vessels seen to report through both systems, LRIT provided
marginally more vessel-days worth of reporting (81% to 88%) than AIS-viaSatellite (76% to 80%). This suggests that, for those vessel equipped with both systems, LRIT is marginally more consistent than AIS-via-Satellite.
The exploratory analysis identified a number of factors that may explain differences between the systems. These included:
- Vessels with a flag-state of registration of Indonesia, New Zealand, Panama, Papua New Guinea, Thailand, Tuvalu, Vietnam or Malaysia were not observed through LRIT during the trial. Note that there may be a number of legitimate reasons, including problems with the LRIT system, which may further explain this.
- Vessels reporting an AIS vessel-type of fishing, towing, high speed craft, pilot vessel, search and rescue, or law enforcement were also not observed through LRIT during the trial.
- A flag-state/vessel-type analysis additionally showed that no Australian/Passenger vessel or Japan/Fishing vessel groups were observed through LRIT.
These findings should be accompanied by a degree of uncertainty that arises from:
- The ‘below-par’ operational status of both systems during the trial period.
This includes indications that there were a reduced number of operational
satellite in the AIS-via-Satellite constellation, and the observation that a
number of flag-states may have been experiencing problems with the LRIT system, - The performance of the AIS-via-Satellite system in areas of high vessel density;
- The implementation of LRIT regional polygons; and,
- Anomalies in the data sets.
In light of this uncertainty, and given the scope in the results for further analysis a
number of suggestions have been made including:
- Repeating the analysis over a longer trial duration and when the systems are ‘fully-functional’;
- Investigating reporting system’s observations versus vessel gross tonnage. Gross tonnage data has not been available for this analysis but if used in the future would provide the means by which to isolate more precisely the group of vessels that are required to report through both AIS and LRIT, and,
- Closer inspection of the vessel population that were observed to be reporting through one system only.
This study provides a number of contributions that may be of relevance to AMSA, the ADO, or to Australian whole-of-government groups with an interest in maritime domain awareness. In particular these findings may act as a guide when considering further analysis for MDA investment options.