A few days ago I was piloting a container ship up the Fraser River (west coast Canada near Vancouver) when I observed on my AIS display a tug heading my way on a reciprocal course at about 10 knots. The problem was that the vessel was not there! Not on radar, visual nor in the VTS system.

Later that day I called the company who own and operate the tug and learned that the tug had been in Allied Shipyard(north vancouver) for the last week. Allied Shipyard is 11 NM as the crow flies from where I observed the AIS target.

So there where two errors:

• Error in position of 11 NM
• Error in course and speed. Actual co&sp Zero at shipyard,,,AIS indicated course made good about 250 deg true..Speed about 10 Kts.

The operators of the tug had no explanation for this anomaly and I have never heard of errors of this magnitude before. If anyone has any theories on this one, I'd love to hear them.

If errors such as this can occur then I have to question the value and credibility of the system.

Feedback

I read with interest the above reported problem (from a Pilot taking a container ship up the Fraser River and who saw a tug on a reciprocal course on his AIS, when the tug was actually in a dockyard about 11 miles away). I would humbly suggest the following possible explanations:

1. The AIS display that the Pilot was using may not have been properly interpreting the "invalid position"  or "invalid course and speed" bits in the AIS transmissions from the Tug. If these bits are not interpreted properly by the receiving AIS display then the vessel concerned can "appear" miles away from where she actually is. Transmissions from Tugs in particular seem to be affected by this on a regular basis and my guess is that it is because of the nature of their duties, as their GPS antennas can become shielded due to the close proximity of high sided vessels, dock walls and other obstructions etc.
   
   
2. Not particularly likely, but many GPS receivers have a simulator mode and it might just have been that a technician was using simulator mode when servicing equipment on the Tug.
   Not sure about this, as I would hope that the AIS on the Tug would alarm if this was the case.
   
   
3. Not particularly likely,  but the AIS specification does include provision for re-transmission of "pseudo targets" by a suitably equipped VTS base station.

Recent experience suggests that some don’t understand the data on the MKD and think ship size is shown rather than position of antenna.

My ship was entering a port which use an own portable computer navigation system based on DGPS and AIS info that keeps them in the middle of the channels and can predict the ETA of opposing traffic. They either use an own antenna that they set up in the centre line of the ship, or plug in to our ais connection. In this case the latter.

While sailing into port, the pilot saw on his system that another ship was still at the berth, although a security call, passing a certain point, had been made. He called his colleague on the VHF only to find out that the ship was indeed underway. Double checking the computer he saw that the ship/AIS was still docked. Confusion galore, negative things were said about our AIS connection, AIS in general, etc etc. On my assurance that my ships systems worked, we contacted the other ship again. There the pilot got the captain up on the bridge (who for some incomprehensible reason was sitting in his cabin) and this captain advised that he had left his AIS system behind in the drydock that he was coming from for repairs. Obviously the shipyard had set up the AIS on a test bench and was happily transmitting away. Not dangerous, but very peculiar to see, with a ship on the way and the AIS locator staying behind.

While sailing further into the same port, we came closer to the dock and saw a bulker docked at the dock. However his AIS was three (!!) cables to the south of him, right at the municipal airport. The pilot passed a message on about this so hopefully they corrected the settings, before they sailed. When docked, such an anomaly is not an issue, however an AIS 3 cables off when approaching a pilot station or VTSS cross roads could cause issues. It gave me the chance to teach the navigators a valuable lesson, NOT to use AIS for anti collision, but what it is meant for, an information system.

Problem encountered:

Our Company tankers have been using AIS coupled with and displayed on electronic charting driven by DGPS for a few years now. My strongly held opinion is that the combination of these two systems give the biggest step forward in navigational safety since the introduction of radar. Hopefully, the progress of ensuring worldwide coverage for quality electronic charts can be accelerated.

Nevertheless, we have just recently noted that one of the AIS models fitted to our ships has the ability to use own ship speed data input from both doppler log and DGPS. But, of concern to us is the apparent fact that the AIS defaults to only doppler speed input whenever the doppler log engages ground tracking mode. The process selecting speed data input source is automatic and beyond any control of the user. Consequently, being aware that doppler speed data is frequently unreliable and can often give rise to large errors for transient periods arising from natural operating conditions we have concern that the design of this particular model is flawed as there are obvious adverse implications for traffic management and collision avoidance. This is now under investigation by both our marine, technical sections and the equipment maker.

– Reported by Shipmaster

Technical response:

The AIS Class A standard does not cover this issue. If it did it could be in danger of overriding IMO decisions concerning carriage requirements. Although the quoted example is arguably not the best possible design, it is compliant to IMO requirements.

The minimum (AIS) requirement is for manufacturers to provide the digital interface to understand the VBW message, which is the IMO required Doppler log message. In addition, VTG, OSD and RMC can be optionally used. VTG is the common message from GNSS equipment giving SOG and COG.

The manufacturer has decided that if ground tracking data from VBW is available it will always take the data from that message, in preference to data from the external GNSS (VTG message). This is not a stated requirement of the AIS standard but nor does it prevent this..

SOLAS requires the carriage of "A speed and distance measuring device, or other means, to indicate speed and distance over the ground in the forward and athwartships direction" on vessels above 50,000 gross tonnage. The VBW message transfers this data to other systems.Therefore, there is a good argument to say SOG should come from a type approved SDME, (rather than GNSS), if it is available. This is what the manufacturer has probably assumed.

The manufacturer should perhaps incorporate more set-up options to ensure that this can optionally be prevented from happening, if this is what most users want.

Nevertheless, IMO does need to give some attention to SOG from GNSS as it is becoming very reliable, particularly when using differential systems and is probably more accurate/reliable than Doppler systems in most circumstances.

Problem encountered:

I observed a constant off-set of AIS position on a vessel I recently piloted. The error remained constant throughout the port transit. Whilst on-board, I managed to identify where the problem was originating. The vessel was equipped with two GPS receivers but they were not providing the same position. Checking with the chart one of the units was providing an accurate position but the other was giving the same erroneous position as the AIS unit. Checking the “own data” function of the AIS confirmed that the position used by this AIS unit was from the erroneous GPS. Unfortunately the Captain spoke very poor English but he indicated that the AIS could not be re-programmed to use the correct GPS unit. I was very puzzled by all this because I always understood that the AIS units had their own integrated GPS receiver which were independent of the ships main system. I also cannot understand why the position was not only inaccurate but so consistently adrift.
I can only think of two reasons:

• The GPS had an error in its internal clock
• It was a GLONASS rather than GPS unit.

On the same screen there is also a vessel alongside the jetty which is also not in the correct position. This small offset is common and I understand that this may be caused by the fact that most AIS units have an internal GPS which is a basic low grade unit rather than a DGPS unit and it is therefore prone to normal GPS signal error. Would this be a correct assumption?

Technical response:

There are several reasons why a GPS may be giving an erroneous error. The most common problem is on older equipment, where it has been set to a datum of other than WGS84. On newer GPS equipment the messages between the GPS and AIS would make the AIS alarm if the GPS was not set to WGS84. I suspect that it wasn't a Glonass problem as the error remained constant. Because Glonass does not have a full constellation of satellites, errors can occur but the magnitude of these errors will change fairly significantly over, say, an hour.

If there are two GNSS (eg GPS) on board one should be designated the primary and the other the secondary. The primary should normally be used unless there is a fault. On some AIS equipment selection of the primary or secondary external GNSS is possible. Most only have provision for one external GNSS. If this is faulty the AIS should be set to the internal GNSS. The Master should have set the AIS to its internal GNSS. Unfortunately, most users have not been properly trained in AIS.

Obviously, it is good practice to use the same positional information for the navigation of the ship and to broadcast via AIS, whenever possible.

In US waters the Coast Guard would take aggressive action against a vessel with such gross inaccuracies in its broadcast position.

With regards to the vessel alongside, varying offsets of up to about 15 metres can be observed when DGPS is not being used. If there is a constant small offset it could be due to an incorrect datum being set in the external GNSS or, perhaps more probably, that the GNSS antenna coordinates have not been set up correctly within the AIS.

In my recent watch cycle I have encountered the following issues:

1. Several vessels heading digitizer (feed to the AIS) is either off or not synchronized
   
2. Static data is incorrect ( in this particular case, no ship dimensions)
   
3. Pilot plug is not wired to the AIS junction box
   
4. No gyro heading feed to AIS
   
5. Obvious lack of crew training with the AIS unit

... a Pilot

1. Installation and Initiation - Several units that I have observed have marginal installations (poor antenna placement, which compromises the range and azimuth of reception).  I have also noted, but with less frequency now, improper static data (e.g. ship dimensions and antenna location)  This can be a minor error or more significant which can lead to target separation on an ARPA display, even at a close range.
   
   
2. Training - It appears that there is little to no training in the operation of AIS units, regardless of whether they are MKDs or fully integrated into an ECDIS or ARPA.  Mariners are self taught.  There are limitations to the AIS tool and these need to be stressed along with its capabilities.

Pilot 

ABSTRACT. This research explores the attitude of Shipmasters towards the use of AIS. It is an attitudinal survey directed at the Master of Merchant Ships exploring their views on a recently introduced piece of shipboard equipment; the Automatic Identification System (AIS). This equipment permits the user to see detailed information of other AIS fitted ships, including its name, position, course, speed and cargo carried. Legislation requires the majority of ships to be equipped with AIS by 31 December 2004.
A convenience sample of 240 ships was used with Masters asked to respond to 10 questions on a 5-point Likert scale. Key findings of this research show that 77% of Masters responding are concerned about the security implications of AIS with 19% switching their AIS off when transiting areas of known piracy. The research also quantified the Master’s understanding of his independence with respect to switching ON and OFF the AIS. Issues of individual training and equipment integration are also examined. Qualitative research highlights the absence of AIS guidance in the round leading to the need for a developed industry wide consensus and improved practical advice to the Master. The research was conducted in June 2004.

Download the full paper (WORD .doc - 214Kb)

I would suggest that the setting of "Sailing" should be changed to something like "Under Sail" to avoid the confusion that seems to exist. Ships that I have pointed this error out to seem suprised that they are in error saying that they are sailing when underway but I am sure that this is not what is meant by the setting "Sailing".

AIS sets are now being fitted in preparation for the July 2004 deadline. My own opinion is that AIS is being introduced in a hasty and flawed manner and my initial experience with using the system confirms that AIS will not provide a safe replacement for radar for anti collision for some considerable time. However it will probably result in vast increase in the use of VHF in developing close quarters situations. On the smaller vessels the ship owners are fitting the basic requirement of a set displaying minimum data on a small alpha-numeric screen. These are dangerous in that they tend not to be fitted professionally and their signal is thus subject to corruption. The input of data into these basic units is fiddly and time consuming. For other vessels tracking them their AIS tracks are prone to irregular errors and we have one regular trader whose vector is frequently reported to be at 90 degrees to the true track when detected by vessels coming up astern!! Another error can originate from the fact that the GPS position transmitted by a vessel originates from the AIS internal GPS. This GPS is only required to be a basic low grade set rather than DGPS so the position displayed on other vessel’s or a ports VTS receiver will therefore only be as accurate as that set. I have noted displacement errors of the AIS target from the radar target of up to two cables. Since there is no requirement for sighting, the units are frequently placed where there is a bit of spare space on the console rather than logically. Distraction of the OOW is inevitable if he wishes to check on another vessel or is required to enter data for port state control or VTS.

The situation is improved when the AIS is integrated with the radar display. I have piloted four vessels with full integration and the best system was fitted to two of them whereby the radar vector was the primary tracking mode and the AIS vector was only activated if the CPA/TCPA ARPA limits were broached by a target or if the target was specifically interrogated by the OOW.

A random and I must admit totally unscientific survey by me has produced the following initial feedback:

• The masters of these vessels with integration all reported that random, inexplicable errors had occurred.
• None of them would trust the information over that of the radar.
• They were all of the opinion that its primary use would be for calling up rogue vessels in anti collision situations. (Not covered by the colregs!)
• Concern that inputting data for port officials could distract an OOW.
• Having input the data, the fact that that data was available to all other AIS users was considered a security risk and a gift to pirates/terrorists.
• Concern that the right of the master to switch his AIS off had been removed.
• None of them had received any formal training for using the system.

With respect to training, I recently attended a course at Warsash and asked about AIS training courses. The colleges are keen to offer such courses but the cost, variety and non standardisation of equipment coupled with the lack of any industry demand or training syllabus renders any such course unlikely at this time.

Whilst any new system is going to suffer from teething problems my own observations have revealed that there are significant problems in both the hardware and software which will take some time to resolve.

MR’S ADVISORY TO SHIPPING No. A-05-2003

202 Lambeth Road London SE1 7LQ +44 (0) 20 7928 1351