This is surprising. I would expect that a North stabilised display would also have a stabilised EBL. Compasses used in stabilising radar have to be fast (>10Hz) so this should be easy enough to achieve, especially if they are already doing the calcs for the EBL readout.
With regard to the article, there are a few issues which are relevant. IMO IEC60954 and IEC60936 give minimum tolerances to be achieved with respect to radar CPA under four standard conditions with 1 and 3 minute trends as follows:
<font class="small">Code:</font><hr /><pre>Condition 1 minute 3 minutes
End on 1.6 miles 0.5 miles
Opening - 0.8 miles
Crossing 1.8 miles 0.7 miles
Overtaking 2.0 miles 0.7 miles</pre><hr />
Although the MAIB report does not say so explicitly, I assume that the Vespucci's master based his estimate of that the Wahkuna would pass 0.8M ahead on the radar ARPA. The ARPA CPA would have shown 2-3 cables to starboard which is well below the minima above. They were erroneously using GPS speed (rather than water speed) as input to theiir ARPA, which would have caused further inaccuracies. They started the second steering motor and went to manual steering, but made no course or speed alteration. BTW, the CPA of 0.8M mentioned in the article was never mentioned in the report.
On Wahkuna they had 2 radar displays: one at the nav station and a second at the helm. Only the helm unit was monitored and I find it difficult to understand how they could interpret the radar as showing the Vespucci passing 1.5M ahead. Even with rudimentary knowledge (and ignoring the EBL), the target range would be seen to be rapidly closing on successive sweeps. In all likelyhood, with collision imminent in the final moments the radar would be swamped by the return and paint an all-round target. I can only assume that they lost all interest in the radar, the one tool which could have prevented the collision.
Food for thought, and interesting to note that their Lokata 121.5MHz beacon failed.
This thread has convinced me that radar installations on leisure yachts should be banned unless the owner has attended a two intensive training course.
I had not realised that the collision detection software in these devices was so rudimentary and disconnected from real world events such as a change in course.
Let me sum up radar. From a highly unstable platform about 16 feet above a yacht a transmitter that is being gyrated all over the place sends out a fat finger beam that at 3 miles is wide enough to hide a fleet of super tankers coming your way.
The computer in the receiver then has to remember the position of the echo's and guess a true bearing somewhere in that fat fingered beam. The radar transmitter then makes leisurely sweep across the horizon. Over a 1 minute period the computer in the yacht's radar will obtain about 5 fuzzy estimated plots of the remote ship whose relative observed angle will have changed very little irrespective of whether the collision risk is high or remote. If the radar is a very expensive model then it should have eliminated most of the gyration effects and additional confusion coursed by variation's in the yacht's heading during the one minute sampling period.
To make use of this dangerously inaccurate information the most able member of the yacht's crew has to disappear down below and start fretting cover a display screen and according to the posts I read, tell the radar system which contacts to observe. It seems it is also necessary to take manual notes of the dangers.
Let’s replay the same few minutes on an AIS equipped yacht.
One both the yacht and remote ship a GPS is sampling its vessel's position at a furious rate and every couple of seconds it can produce a position that is an order of magnitude more accurate than the guestimate that can be obtained from a remote radar transmitter.
Over a 30 second period both GPS's can produce a highly accurate heading and speed. Every 10 secs the big ship will send out this unambiguous information in a sub-second digital VHF squawk.
Then in the yachtsman's pocket a device a little larger than a mobile phone will receive this pin-point accurate information via a bluetooth signal sent from a block box at the chart table. Inside the handheld device, software that is simple enough to have been written as part of a GCSE level computer science project will apply GCSE level Cartesian plotting arithmetic to assess the risk of a collision.
If the CPA is too close the device will vibrate and bleep. All the yachtsman has to do is keep on hand on the helm and with the other hold the pocket AIS device, read a couple of figures and look in the direction of the flashing black line.
In short the owner of the AIS device can behave like a proper mariner.
If you have not seen it before, this http://www.mcga.gov.uk/c4mca/mcga-6-456rep.pdf research report done for the MCA may be of interest. Looks at the association between radar and AIS data in collision avoidance and what should be relied on, but is from the big ship perspective. Is quite long but mostly easy to read.
Obviously, in comparison to the paper, little yacht radars (small screen, tiny antenna, etc) are much less accurate both from the display and user knowledge points of view.
At least radar, with its imperfections, paints a picture. A yacht could just as easily have been run down by a trawler without AIS in which case your handy device wouldn't utter a squeak and it wouldn't just be your teeth which had the blues!
IMHO this is undergilding the lily. Generally the USE of radar is invaluable in collision avoidance given knowledge of the basic use of radar and a good knowledge of collision regs. It is interesting to note that in this case, radar whilst switched on was not being monitored on a systematic basis. Had radar been used properly in this case then it is highly likely that the collision would have been avoided. From what I gather, the use of radar did not contibute to the collision but the lack of use, as well as the incorrect use may well have.
Sorry badly worded I don't do close quarters big stuff in poor vis; however in good vis I will work down to 2 miles with range rings at 1/2 mile on selected targets if plotting and doubtful, the point being that the apparent relative bearing on screen changes more quickly on those settings, not in degrees/minute which is obviously constant, but in distance travelled by target across screen (badly worded) which makes plotting easier, or so it seems to me.
I've just read the article - very interesting.
A lesson I have learnt from flying - especially as I am in the middle of an Instrument rating is that the instruments (i.e radar in a boat) MUST be viewed in the context of your current heading, and to a lesser degree the speed. The situation painted in the CA article - which is very believable - would probably not have occured if the heading was viewed along with the radar picture. I'm not saying the collision wouldn't have occured but as soon as the boat was put back onto the correct heading the picture would have changed dramatically.
Thanks. I did some speed reading. Seems a major theme of the report is how to present divergent plots of the same vessel arriving via AIS and radar, onto a single screen.
I will need to reread the section on why an AIS bearing becomes less reliable at under 1/2 mile, surely with a 10m error at each end of the bearing the issue cannot be that significant.