And fishermen, yachties, sail boaters and jet ski drivers all climb Pilot ladders? Shooting from the hip again……
According to the IATA, the average for fatal accidents involving commercial passenger and cargo aircraft is 6 per annum. Now just to reconcile your issue with the mathematics, let us assume that there is only one Commercial Pilot in the cockpit……that displays a ratio of 6/1.8……as I said……less than half.
One space shuttle uniquely designed to complete that task at enormous expense. Now, let us go back to my earlier screenshot…
This is a small representation of the total number of vessels sailing the globe at any one time. For these vessels to be controlled by AI software they must be suitably designed and compliant. When you look at this picture, that is an enormous replacement task at astronomical cost.
With regard to remote Pilotage…… relying on the vessel’s Master and bridge team is a very scary thought when matters go pear shaped. Sorry to reiterate, but I am speaking from experience in this instance.
There are some eminent scholars and engineers warning us of the potential dangers relating to the development of AI technology.
amazed that some of these incompetent sailors somehow loaded all that cargo, managed and maintained that $100M asset and found your port in the first place.
Sorry to reiterate - but I have seen, and worked with a handful of pilots, just a handful, that I was very scared what would happen if things went pear shaped as well.
There are. Several companies offer Straits Pilotage:
https://www.strasselink.com/pilotage.php
PS> The first Straits Pilotage service was started by Keenan & Co, back in the
mid-1960s. My Father-in-Law worked for Capt. Keenan at the time and did some pilotage, especially for APL ships heading to Port Swettenham (now Port Klang)
He took over the business in the early 1970s.
It feels like you are getting hung up on the shortcomings of current technology and equipment as justification that it can’t change.
Can you specifically indicate to me where I used the word “incompetent”? Very strong language.
Of course there there is not an homogeneous skill level within all Pilots……in my neck of the woods….they have come from either being a Master or Senior Officer. Like all professions, there are different levels of abilities, situational awareness and skill sets.
Perhaps you should watch the Geoffrey Hinton lectures.
you are correct - you did not use the word -
but incompetent is not a stretch of a description of this
thanks - and once again - 99.5 pct of every interaction I have had with pilots has been nothing but capable and professional -
would have been 99.9 except for NOBRA pilots in the 80’s !!!
Remote pilotage and AI replacing pilots are two different issues, but the discussion alternates between the two as if they were the same thing
For me, “remote pilotage” means an actual pilot is still navigating the ship, but the pilot is not physically present. I think this is doable with existing technology. 10-15 years ago I was using my ppi to turn ships in basins and we were all running channels in fog with our head in the radar since the 1980’s. It really doesn’t matter if the radar repeater and ppi are on the ship or in my bedroom. Add video visuals for even more control. In all but the tightest situations that would be perfectly possible In those worst cases I probably would’t be moving around in a ship anyway
AI replacing pilots is a much more difficult problem. People who have never piloted have a difficult time understanding the level of expertise needed to navigate a heavily loaded ship in a narrow channel where the ship is all but dragging its keel through thick soup. There is a long S turn in the Houston ship channel that is quite often navigated using left rudder only on the right hand turns and using right rudder only to make the left hand bends - just by controlling the position of the ship a few yards to right or left of the centerline.
Simulators have been around for 50+ years now, but the mathematical models and algorithms that govern them are based on deep water data. They are totally lost with a bank on just one side, much less a bank on both sides, a bottom clearance of less than half a meter and large passing vessels.
A car is an infinitely more simple problem and I’m not seeing much progress there.
Give me Ai that will rout me accurately to a customer service person on the phone. Give me enough AI that will just operate a traffic light so I’m not sitting stopped at 3am with no traffic in sight. How hard can that be? But here we are discussing driving tankers full of gasoline when we can’t even write an algorithm to find a movie I’d like to watch on Netflix. Let’s do the traffic lights first, then the tankers.
I was second mate at the time and my first trip to Europe so my memory is a little hazy but I do recall one time in N. Europe being guided by a pilot remotely.
From here (June 2020): Remote pilotage – perspective and risks to consider
We are also aware that the practice has been used for many years, particularly in north-western European ports, when adverse weather conditions outside of specific ports prevent the efficient movement of shipping over a sustained period.
“The practice” that “has been used from many years” is remote pilotage.
Good point on separating the issues. 100 percent the “remote pilot” is much more doable than the “AI” pilot.
But I don’t think those are mutually exclusive. The perception of AI being a hands off, computer in full control, idea is not really how this or other uses of AI will be used. AI is a tool, just like lots of tools out there - many, like the PPI you mentioned - which are so rapidly now part of our lives.
I have little doubt that driving a deep tanker down a narrow channel is not multiple more times a greater challenge than landing a plane, or a driverless car in a city. At the end of the day it is a 2 dimensional problem, with some - probably large number of variables, and situational inputs. AI will be able to do this - at some point.
As talented as you and all the HSC pilots are, you were not born with that innately, you were trained to be competent, and experience made you an expert. AI, at some point, will be able to do much of that as well.
Now what I see in my crystal ball well down the road are command centers, manned by local experts, that assist vessels in transits in some areas. AI will be a tool in these centers, that will be monitored by humans. Is this a drastic change in the number of pilots needed - probably.
There is long long line of folks who didn’t think the next technological advance would impact them, wouldn’t work, wasn’t possible. To date I don’t think any of them have been proven right.
I think you are right, on both points.
We are not talking about ships docking, or even moving in and out of port, or in narrow canals and fairways, without a pilot onboard. (At least not yet)
The Danelec trial is about ships transiting the Belts or the Sound and limited to ships of a certain technical standard and draft (-13m.)
This is fully feasible in well charted territorial waters, with good radar coverage and VTS in force.
As said several times here; it is already widely used to guide ships into smooth waters when the weather does not allow boarding in open waters and v.v.
“it is a 2 dimensional problem, with some - probably large number of variables, and situational inputs. AI will be able to do this - at some point.“
Only 2 dimensions? I’m counting 4, at least. 
No doubt - eventually. By the time AI gets that good we might have solved the gravity issue and ships will be airborne.
Have a good weekend.
String theory !!
Time is usually the fourth dimension. String theory and quantum entanglement are lurking in the pile after that.
man - I have still never won an argument with a pilot !! 
Not an argument. Just banter. 
AI will probably eventually replace humans at everything and cause a global collapse of society, humans were not designed to sit around collecting AI welfare checks. That day is not today.
Here is a fundamental question: Do pilots provide local knowledge or make up for lack of ship handling skills too?
Reading about the loss of the pilot boat Can Do, it seemed she did her job by meeting a ship and then leading them into harbor, the assumption was that the ship’s crew could manage to follow the boat. Was/is that unusual?
Yes.
