[Varley]

At the risk of boring those who have seen the tales on other threads (I cannot remember them all, no true bore can) I can use them to illustrate some of my prejudices. Best of all was during LNG seatrials off DSME. I suppose we could think of her as some form of hybrid between auxiliaries fully integrated or replaced('duplicated') by IAS and those simply remotely controlling the maker's kit from the IAS. A steamer, the burner management and combustion control done by Kongsberg but the turbines/manoeuvring controls very much the makers (Kawasaki).

As an example of how contactors should not be joined in the building process I offer under. One good process model is the one described "V" shaped. Well presented in a presentation on the (mis)application of IAS by LRS's Bernard Twomey, always a good man to hear.

(Before E-S jumps in that is not a plump round shaped map but one where the planning and drawings, here I would include software, are completed during the first downward stroke of the "V" with the casting into ship starting with the keel laying at the bottom ending in delivery at the top of the upward stroke).


The Kongsberg team of about five with a consultant from HO were plugged into the network with laptops, at a table arranged in the MCR. They were not simply adjusting parameters but they were 'coding' on the fly. Perhaps not unusual when sliding down the leading edge of V but we were underway on sea trials.

I have a bug bear about the lifetime of high populations of silicon WRT single failure prone essential equipment. Without claiming that it will fail in ten years I do claim that one must be budget ready to replace it if it does. Many experiences of being forced to repair kit in which the remaining semiconductor population is on another rising stroke, that of the bathtube curve, has taught me that to do so results in a higher cost and greater down time. My boss is amused to know that my doorbell defies my crystal ball. It is based on a PLC that I bought to demonstrate to colleagues what one was, in the late 80s. One swallow doth not a Summer make. I am still, as always, right.

The replacement electronics itself will be cheap. In ten years, though, it will be different and if the interface between the original and the 'ship' is not well documented then the replacement will involve engineering a new one. That will be expensive. That documentation is simply not being done (perhaps LRS have managed some improvements there I have not spoken to Bernard for some time).

The consultant on the trials kindly gave me an introduction to the kit and listened to me unravelling my spiel to a new audience. His honest agreement I now quote at the drop of a hat.

"In ten years time this system will not only be expensive to maintain, it will be fabulously expensive to maintain".

Were the fatigue life of commercial tonnage to be restricted to 10 years I would be relaxed but that isn't the standard and I forecast that we will be scrapping because of the 'fatigue' life of her silicon.

A model that I would have liked to see adopted (it was available and not too dissimilar to the Decca Isis I 'did' at Saudi Shields in the 70s) was to have a discrete alarm system like the Autronica KM series or Malling with whatever computer enhancements you could dream up claggted on the back of it.

On hearing mention of Autronica and KM the consultant told me that Kongsberg had bough Autronica and still produced a KM variant. "For out clients operating in areas where it is difficult for us to offer support". Where I wonder is more difficult than mid-pacific?

(How old was Longbow? She went to scrap with her original Autronica KM2 (?) operating fine(ish - complete drawings would have been helpful there too!).

At the specification stage of a combined LNG/Gas production vessel I included a line or two to counter this behaviour. My oppo. in the client's office was from the oil and gas production industry and I learned that there, with the same kit, they would expect it to have arrived at the 'project' having been fully tested back to back with a simulator with the expectation of a few parameter setpoint adjustments only to be made to an otherwise fully operational control and monitoring system.

I don't know about the RN but certainly merchant shipbuilding is not well served by these contractors (I am sure Kongsberg is not the only one but my experience with others is slight).

They serve public transport, nuclear and other areas under more public scrutiny, they know well how to deliver but we obviously don't present a sufficiently lucrative market.

(I would have revelled in it to have had such kit as my preserve at sea, great toys to play with and to enjoy the kudos and status of being the shaman in charge. The picture is very different from behind the office desk).

One of the most interesting little problems in our long and well watered run down to redundancy was a systems fault that occurred on a Kongsberg Datachief. It occurred after twelve years in service but was then permanent - another story I have told elsewhere but I will wait for your eyes to unglaze before I launch into that again.

(I also dislike Kongsberg's approach to redundant power supplies. The only time a UPS saves one from whatever is when one's computer can avoid the whatever without a propulsion/power plant to respond to it. Otherwise their failure risks killing the computer while the main supply is healthy and, of course, is also more likely to reveal a hidden failure to UPS when it is required to switch over).

I really do need to let you recover.