Kneejerk Legislation and Cosco Busan bandwagonning

Dennis Bryant at Holland & Knight clearly shares our exasperation at bandwagonning politicians climbing aboard the Cosco Busan. Commenting on a bill introduced by that global maritime expert Senator Feinstein, he says ‘The bill provides, among other things, that during a condition of enhanced danger (which includes dangerously low visibility, whatever that is) the USCG Sector Commander “shall assume direct authority over all vessels within the area . . . to ensure the safe navigation of dangerous waterways.”This is a truly misguided approach that will result in more problems than it expects to solve. ‘

What one wonders is whether that sector commander will be subject to criminal charges, face imprisonment and lose his career should an incident occur on his watch? Will he face the same legal risks as the pilot and ship officers aboard the Cosco Busan. Will the Coast Guard be fined millions of dollars for any pollution that occurs as a result? After all, if he, or she, is to assume direct authority over a vessel then he or she should be liable, as should the Coast Guard, so it’s little surprise that the Coast Guard itself really doesn’t want that particular hot potato.

In fact, there was little the VTS could have done at the time of the Cosco Busan incident. Its obsolete equipment had a time-lag between a ship making a manouvre and that manouvre being seen on the VTS screen, The equipment was obsolete because politicians had voted down funds for its upgrading.

Even modern equipment is not ‘real time’, there is always a lag which can be significant when dealing with fast-moving vessels.

VTS is not air traffic control. Every craft in the air is subject to air traffic control but the waters are filled with small vessels with little or no radar signature, without AIS, without radios, which can, and do interfere with safe navigation but will not be seen by the VTS.

Dangerous situations can arise without warning. At that moment, those on the bridge must remain focussed and undistracted, which they can’t be if they required to take orders from someone who isn’t on the bridge and doesn’t know the immediate situation. What will ensure will be a negotiation that will interfere with the bridge team’s situational awareness, particularly if the Coast Guard officer behaves with the typical abrasiveness of the breed.

One must, however, beware of the same sort of knee-jerk reaction that is influencing Feinstein’s bid for maritime glory. One must question whether direct VTS control is necessary or advisable in waters with mandatory pilotage, otherwise, why have a pilot anyway? Again, the issue of liability has to be settled.

There may well be an argument for greater VTS control, particularly in areas of high traffic density, and given the decreasing experience of many ship’s officers today, but ill thought-out legislation isn’t the way to go.

10 Responses to Kneejerk Legislation and Cosco Busan bandwagonning

  1. David Hindin says:

    “In fact, there was little the VTS could have done at the time of the Cosco Busan incident. Its obsolete equipment had a time-lag between a ship making a manouvre and that manouvre being seen on the VTS screen, The equipment was obsolete because politicians had voted down funds for its upgrading.”
    Please provide a source for the “fact” of the time delay.
    How much time delay is being claimed?

    Thanks

  2. Bob Couttie says:

    Allision of the M/V COSCO BUSAN With the San Francisco-Oakland Bay Bridge,
    Departmenmt of Homeland Security, Office of the Inspector General,
    April 9, 2008:

    “The VTS watchstanders on duty before the accident could not have taken any additional action that would have prevented the casualty. A time lag between a maneuver that a vessel is executing and the vessel movement data displayed in the VTS operations center precluded this possibility. For example, the VTS contacted the
    pilot onboard the M/V COSCO BUSAN pilot at 8:27 a.m. (see Table 1) to advise him that the VTS center’s equipment indicated that the vessel was on a 235degree
    heading; the pilot informed the VTS operations center that the vessel was executing a turn and that he was steering 280 degrees. By the time the VTS watchstanders recognized that the M/V COSCO BUSAN appeared to be out of position to pass between the Delta and Echo columns of the San Francisco-Oakland Bay Bridge, the vessel had already started to execute its turn towards the bridge column.”

  3. David Hindin says:

    Your reply does not really address or quantify any significant technical delay between the VTS situation display and the real AIS data,. The example you gave is interesting in its own right: “VTS contacted the pilot onboard the M/V COSCO BUSAN pilot at 8:27 a.m. (see Table 1) to advise him that the VTS center’s equipment indicated that the vessel was on a 235 degree heading, The pilot informed the VTS operations center that the vessel was executing a turn and that he was steering 280 degrees”
    I refer you to the NTSB Docket Management System for the CB incident
    http://www.ntsb.gov/Events/2008/San-Francisco-Bay-CA/Exhibits/default.htm
    Document 389887 “Voyage Data Recorder (VDR) Audio Transcript”
    http://www.ntsb.gov/Events/2008/San-Francisco-Bay-CA/Exhibits/389887.pdf
    On page 133 of the referenced document: 08:27:48 VTS “unit romeo traffic uh AIS shows you on two three five heading” (which is the transmission you have quoted). The companion combined radar /navigation display shows that the ship was actually on a course of 235 at that time and a heading of 261.5. Telling a captain that his heading was two three five when he in fact was on a course of two three five could be construed to constitute an error by VTS affecting the pilots already compromised situational awareness. (It would also allow some misguided claim of delay) There is some question as to whether the San Francisco VTS display system actually even presents heading to the operator (which may make your point about obsolete equipment). I can’t provide a firm reference to my last point without replaying the NTSB hearing or waiting for the formal transcript. I can tell you that the AIS data tagged 16:27:37 Z (08:27:37 PST) showed Cosco Busan:SOG: 11.3 kt HDG: 258°T COG: 235.1°T.”. Earlier at 16:26:58 Z (08:26:58 PST) AIS for Cosco Busan shows SOG: 10.8 kt HDG: 243°T COG: 238.7°T. At that time, the ship was on a course generally parallel with the bridge and the bow was at or past the intended span center. The allision happened more than two minutes later. The mixup in the VTS call between heading and course could not have been helpful in averting the allision. An earlier correct call might have. I stand by my original claim that no substantial delay between the available AIS data and the VTS display has been demonstrated. For a look at a time tagged AI display of the event, I recommend (from the gCaptain site):
    [video src="http://s3.amazonaws.com/gcaptain-s3/maritime/video/dec07/CB_Thru_Span_2_dbl-Desktop.m4v" /]

    Thanks

  4. Bob Couttie says:

    A number of technical issues arose during the investigation which make it a little difficult to synchronise various bits of data: “The San Francisco VTS operations center has the capability to record radio communications and automatic identification system data transmitted by vessels in its operating area. However, on November 7, 2007, the VTS and the M/V COSCO BUSAN transmissions were not synchronized to the same time mark. According to the Coast Guard, it is suspected that the timing signals for the recorded data were not resynchronized following maintenance of the San Francisco VTS equipment approximately a month prior to the allision. The failure to synchronize recorded data impairs the ability of investigators and other interested parties to accurately recreate the chain of events leading up to and following an incident.” (DHS,OIG Report).

    The animation is based on AIS data received by a third party, http://www.boatingsf.com after the fact, positions are approximate and do not necessarily reflect what was on the screen in the VTS.

    The following statements were made by Mark Perez, VTS supervisor at the time of the incident:
    “So the AIS had not in our opinion, at
    10 least in my opinion, the AIS had not quite caught up with the
    11 radar track. So —
    12 Q. Okay.
    13 A. — because AIS, depending on the speed of the ship and
    14 its rate of turn will only update every so often, and typically
    15 it’s every six seconds in a, in a turn.
    16 Q. Okay. And that — where do you draw six seconds from?
    17 A. That was from the information we were given to us on
    18 training —
    19 Q. Okay.
    20 A. — when AIS first came online.
    21 Q. Okay. That was from an independent contractor or —
    22 A. It was from our Training Coordinator, Scott Humphrey.”

    and:
    13 A. — the —
    14 Q. — as far —
    15 (Simultaneous comments.)
    16 Q. — spike or a vary or anything like that?
    17 A. We have — each vessel will have a track leader, which
    18 will basically be a — basically DR, whenever the vessel will be
    19 and to a certain amount of time depending what the chart’s set up
    20 for. Our Central Bay chart is set up for three minutes. So we
    21 have an idea in three minutes where that vessel will be based on
    22 that track leader.
    23 Q. So in a sharp turn, does that become unreliable or —
    24 A. It’s a little disputed because the, because the data’s
    25 coming in. I mean it’s every six seconds, the data is sent out.

    1 How our system actually processes all that data, I’m not 100
    2 percent sure. Does it grab it all? We don’t physically see a, a
    3 course displayed. We have to open up — either open up the, the
    4 transit card, which will give us a, a static display of the
    5 course, so it won’t update while that card’s open; or we can go to
    6 what we call a sector summary, which will show — an AIS updating.
    7 But it doesn’t correlate with the — it’s close to the radar
    8 track, but not the actual track.

    NTSB docket 389165.

    One might, I suppose, regard a six-second delay as insignificant, but it means that the VTS is not seeing realtime information at a critical moment.

    Indeed, as you point out, the VTS was in error when it advised the pilot that he was on a heading of 235 as opposed to a course, when the heading was actually 261.5. VTS believed the disparity between what was on their screen and the pilots claim to be steering 280 was due to the time lag in the system.

  5. David Hindin says:

    Thanks for taking the time to reply. I have enjoyed this dialog.
    You had noted “The animation is based on AIS data received by a third party, http://www.boatingsf.com after the fact, positions are approximate and do not necessarily reflect what was on the screen in the VTS.” Very true. The AIS data is believed correct as fair as it was available (one channel) and as processed to recover time tags. The method of AIS display at VTS is certainly in issue. Relying on a three minute leader (as you noted above) might have worked in the case of EXXON VALDEZ and its encounter with Bligh Reef but is certainly not appropriate for close quarters like the San Francisco Bay in the vicintiy of the Port of Oakland. Basically all data available to the VTS is sampled: The radar data by the rotation of the antenna and AIS data by an algorithm which depends on the speed and status of the ship.
    i will accept your argument that sampling implies delay. An extract from the ITU RECOMMENDATION ITU-R M.1371-3 “Technical characteristics for a universal shipborne automatici dentification system using time division multiple access in the
    VHF maritime mobile band”
    TABLE 1a Class A shipborne mobile equipment reporting intervals
    (NOTE 1 These values have been chosen to minimize unnecessary loading of the radio channels while maintaining compliance within the IMO AIS performance standards)

    Presented (extract) as Ship’s dynamic conditions and Nominal reporting interval

    Ship 0-14 knots 10 secs
    Ship 0-14 knots and changing course 3 1/3 secs
    Ship 14-23 knots 6 secs

    Under the dynamic conditions present for CB she could have been reporting at intervals as short as every 3 1/3 second. and as long as every ten seconds. At ten knots, a ship travels about a thousand feet ( a little more than the CB LOA) in a minute. Using an average update rate of every six seconds, new data is available for every tenth of a ship length advance. I do not believe that this sampling interval is outrageously long to adequately represent the ships position and status for collision avoidance operations, Would I want to remotely “con” the ship at this sample rate? Probably not recommended, but not impossible. What is in question is how VTS acquired and displayed the data. I noted the three minute line of advance argument above. I am also led to believe that VTS San Francisco doe not display heading to the operator, which can result in the confusion also noted above. How is AIS data integrated into the VTS displays. This question does not appear to be adequately addressed in the documents available at the NTSB Docket site,.

    Again I appreciate the opportunity to engage in this dialogue.

  6. Nice article, I agree Feinstein’s legislation is shotting from the hip. Everyone loves solutions that are easy to explain to the public, for example double bottoms – keep up the good work

  7. David Hindin says:

    My final thoughts on this thread,
    Had the COSCO BUSAN pilot used an available personal pilot unit (PPU) connected to the CB AIS set through the required “pilots port/plug” he could have seen the same presentation shown in the animation referenced above (and the frames prior) and we simply would not be having this discussion. Immediate situational awareness, no confusion and no allision.

  8. David Hindin I think you are likely correct, that in this case, had the pilot had his own PPU the accident might not had happened. However , long term, I believe that to achieve the highest degree of safety in say, S.F. bay, a three-legged Pilot(with PPU/)-Bridge Crew-VTS system would be optimal. Each has it’s own strength and weakness and a trifecta system so to speak would have overlapping areas which would be safer then each element alone. Feinstein’s legislation on the other hand seeks to simply shift responsibility ship side to shore side, which is not the way to go as Bob points out.

    A single element any system can fail, it shouldn’t pull down the rest of the system with it.

  9. Excellent technical discussion. several factors are relevant: !. VTS sits on hill in the middle of the bay The line of sight dfrom VTS to CB ranges from 1640 yda to D tower, 984 to CPA with CB and 3116 yards to outer harbor Buoy #4. Assuume the VTS is being observed on one unchanged range scale and by one operator, alert/recognition time is probably several antenna rotations.The PRR will dictate the quality of presentation that is constantly changing 2. A medicated person directing the navigation and movements of the vessel (pilot) with a history of navigation errors etc., is not your average bar pilot ( I hope) therefore one should view when and where does an indication of inappropriate action appear. AND, who should notice it ? We have here the breaking of the myth that pilots are infallable. 3. The recorded data from AIS, VTS radar and ships recording system use different terms and definitions, Coordinaintg is necessary as each appears to be reporting the same thing differently.(I am still confused, but enkightened).
    Comment: ,In limited visibility, less than a ship length, the conning person is best stationed at the radar in open waters. The Oakland Bar Channel for tug oriented pilots is like the ocean,–for seagoing oriented pilots it is congested, therefore if concerned about a collision or allision it is one’s perspective as to how to determine immediate danger. However if confused and in doubt, as I feel Cota was, his rational response was inoperative and he reacted with several rudder orders and a speed change with different and confused solutions in mine. The last thng he needed was more guidance and electronic data input. At fault yes, but for reasons of confusion and under the infludence of medication and electronic inpt stimulation. and possibly lacking support from an unprepared BRM An old pilotage rule: if in doubt, mumble, someone may do something right. Refer : THE CARE AND FEEDING OF SHIP’S PILOT; U.S. Naval Institute Proceedings, Aug 1993. by guess who? JGD

  10. Bob Couttie says:

    The latest news is that the US DOJ is to press further criminal charges against John Cota. I’m sure we all appreciate their enthusiasm for clearing the seas of pilots and the contribution they are making to the safety of the seas. Or maybe it was just a slow day at the water cooler.

    I would point out that Senator Feinstein’s research assistants haven’t done their job: The VTS operators already have the authority to direct vessels if a situation warrants. Also, I’m not particularly of the view that double hulls are quite the solution they’re painted, but that’s a separate issue.

    Nobody, I think would dispute that both VTS, and those on the bridge of the Cosco Busan lacked situational awareness. However, situational awareness is rather like good sex on two counts: A) everybody knows what it is but everybody has a different definition and B) You don’t know you haven’t got it until too late (Google TADMUS for much good work on the issue, situational awareness, that is, not sex). Loss of situational awareness is a symptom, not a cause.

    Notably missing from Feinstein’s legislation are the two most critical elements, Bridge Resource Management, BRM, and Bridge Team Management, BTM (BTM being a subset of BRM), which are far more important issues that whether Pilot John Cota had a DUI a decade ago and vital to situational awareness and good judgement.

    So far, I’ve seen nothing in any reports, such as the DHS-OIG report, or the NTSB dockets, to suggest that revision of my initial estimate of poor BRM and BTM is justified.

    John Cota’s trial, which will also implicitly be the trial of the officers and crew of the Cosco Busan, will be a circus, although ‘Grande Guignole’ theatre would be a more appropriate term. The aversarial process of criminal court procedures in the United States is ill-equipped to provide solutions. Even should Coita be found guilty and fined it will no more improve maritime safety than the hanging of ten-year olds in 19th century London for stealing bread solved poverty in Shoreditch.

    The only beneficiaries will be lawyers seeking to pay the mortgage on their next BMW or holiday home in the Seychelles, or alimony, and opportunist politicians whose clarity of vision echoes John Cota’s view of the Delta-Echo span..

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: