Piloting, Autonomous Vessels, AI, and the coffee making machine

Text and Photos by Captain Ricardo Caballero, Panama Canal Pilot

I am not a computer savvy. My knowledge in programming and robotics and those sort of things is nil. I get lost in the sea of social media and easily entangle myself in the web. All I have done for the last 25 years or so is to pilot ships through the Panama Canal. However, during the last couple of years I have done my best to catch up with technology, since it has enhanced our possibilities and improved safety in our field. But still, I have to admit that I am way behind the new guys in this important issue.

Artificial Intelligence, AI, is challenging humanity in every possible way. It is changing the way most tasks are carry out these days. It is making life easier for some and, at the same time, harsh for others.

Decades ago workers began to be slowly replaced by machines. The rate of replacement has been accelarated by innovation and new technology. In his book “21 lessons for the 21st Century, historian Yuval Noah Harari predicts that within the next 50 years jobs such as truck drivers, taxi drivers and the like, will be lost to computer algorithms.

But it is not only on the highways that humans will lose their jobs to computers. Already in the open seas the battle has long started. ARPA ( Automated Radar Plotting Aid) emerged to ease the duties of sailors when traffic and weather were increasing the rate of maritime accidents. It started with Radars which evolved into ARPA. Soon the GPS was released and incorporated to conform an integrated system that allows the officer on watch (OOW) to ascertain and predict situations at a glance. The further development of these integrated systems and the advance in bridge design might have led the industry to think of reducing the manning on their fleet. It is true that, by incorporating all this new technology, not only could the Officer on watch take a more accurate decision by the information provided by the system, but the equipment itself would show him the best course of action to be taken.

All these did away with many of the, often tedious and time consuming process, of manually plotting the targets on a radar sheet to avoid a collision. Later, the introduction of electronic charts enhanced even more the availability of real time information for the watch standing officer eliminating the need of plotting the ship’s position on a paper chart that was probably not properly updated. Great!

The reliability of this kind of equipment, paved the road for IMO to eventually make it compulsory aboard ships, but in no way to replace the bridge team members. But for others aboard, technology was not as gentle as it was with the members of the deck department.

As an example, the GMDSS, or Global Maritime Distress and Safety System, put an end to the use of Radio Officers. The system ensures rapid alerting of shore-based rescue and communications authorities in the event of an emergency. There is no need to have an expert on communications aboard, since each one of the officers is trained to operate the GMDSS.

In the Galley department the Chief Cook was left alone as only “Cook” and his staff was reduced to a single steward. It might seem as a minor change, that only impacted the quality of the service they provided to the officers aboard and the workload of galley personnel, but it happened also because technology found its way into the galley.

The use of remote monitoring systems for the main engine and auxiliary equipment gave room to what is now known as “unattended engines”. This allows the engineers to optimize their time, say on performing maintenance work instead of constantly watching the engine indicators to see if they are working fine. Now they can even have some extra rest time during the night since an alarm in his cabin will alert him of any condition that requires his presence in the engine room. Great!

However, the race for reducing costs by downsizing ships crews is a fact not to be taken lightly. It is a race that started by promoting the development and improvement of new technology in order to take safety to the next level. Well, and save big bucks in the meantime. Now, with AI lurking on the horizon what follows, or at least it seems to be the main idea, is the creation of “autonomous vessels”.

From 2015 there have been projects dedicated to the development of autonomous ships. Finland, the United Kingdom, Norway, and China are among the countries that have taken the first steps in that direction. There are already small prototypes built to test the feasibility of self governed ships. And there is already a vessel, the 80m long with a capacity of 140 containers, “Yara Birkeland”, which is expected to be in operation by 2020, that would not only sail (by 2022) with zero crew aboard but that will also be the first zero emissions boat. That is good news for the environment, and probably an indication that the need for sailors is not too far in the distance to come to an end.

Wait. Not too fast. First the shipping industry will have to make the transfer from crewed to autonomous ships. For that to happen, they will have to adapt their ships so they do not require human assistance, or minimum human assistance. Fitting all existing ships with the required equipment for them to become autonomous would be very costly.
Normally, when the shipping industry goes through important changes that include the incorporation of new equipment or the refurbishing of their vessels, a phase out period is required. Changes of these kind cannot be done overnight, not even over year.

Second, new regulatory systems from both countries and IMO will have to accommodate to the new standards. In the case of SOLAS, the Safety of Life at Sea Convention, will most likely have to be rewritten.

The consequences of autonomous ships cruising the oceans will be far fetching. However, I doubt that any of this is going to happen in the near future. The closest to an autonomous vessel that I have had under my command are the latest generation of passengers. These ships are equipped with what is called Dynamic Positioning System, or DP.

DP is a computer-controlled system that automatically maintain a vessel’s position and heading by using its own propellers and thrusters. The system controls the ship by using information provided by sensors for an array of factors such as wind, current, and gyrocompases to then activate its propellers to anticipate from any changes on the ship’s position and heading. Ships with this system do not, under normal circumstances, need to drop anchor. Fantastic! Though I must remind you that only “under normal circumstances”. However, it also has its limitations and needs to be checked by the OOW.

So if today’s sailors can relax that their job is not being jeopardized by Artificial Intelligence and its younger brother “autonomous ships”, in the mid term, how about us Pilots?

I think we are farther away than regular sailors from being tackled by the tentacles of technological innovation. This is not to say that we Pilots do not agree with technology. We cannot disagree with it. Our homes are stacked with technological “gadgets”. In fact we depend on technology everytime we pilot a ship.
The Pilot Portable Unit, PPU, has become our favorite navigational tool next to our transit radio. The success of the expanded canal, with its challenging locks and the handling of the massive ships, its due in part of the technological side of the story. The accumulated experience of the Panama Canal Pilots, pitched together with innovation technology is facilitating the transit of neopanamaxes with almost no incidents.

How AI could replace us?
It could. But again, not in the near future, not even in the not so distant future. I have already mentioned some of the reasons above. We can add probably the most relevant reason to us.

Statistically speaking 80 percent of marine accidents are due to human error. Does this mean that by replacement humans with “intelligent machines” (in our field ) the number will be cut? And if it does, by how much? To answer this questions we would have to classify the accidents and the net impact they have on everything, and everything should be reduced to humans, but at the end, everything seems to be money.

I am not going to dig out recent data related to accident rate in which the consequences of the remaining 20 percent in which there was no human error were catastrophic or not.

In reality, human error is always present. Machines are built by humans, other humans, not us Pilots. And for us, machines are just additional tools that aid fulfilling our work goals. They enhance our senses, they help us to overcome our limitations by providing useful information that sometimes is not even reliable or that our senses are not designed to detect. But they cannot replace us.

Piloting involves the awareness of our five senses, and the way how the resulting information is processed and interpreted. (Sure, smell is also included).

Proprioception is the ability of our brain to make itself aware of the movement of our body. It tells us our position in relation to our surroundings. Without it we become disoriented, lost.

The Pilot’s “proprioception” extends to the structure of the ship under his command. For this to be accomplished the pilot needs to know very well what its surroundings are like and what the characteristics of the ship interact with it. What it is commonly referred to as “local knowledge” spiced with a high level of intuition. It takes several years to fully master this “local knowledge”.

And just to make things clear, local knowledge is not restricted to physical elements. It also involves a comprehensive understanding of the procedures within the pilot’s domain. The way communications are made, the manner in which resources are used, and even the local jargon.

The possibility that a machine built with the ability of carrying all the necessary algorithms to do our job is, of course, there. The probability for that machine to replace a pilot, ignoring that if something goes wrong there would be no human brain to take an alternate route not included in the program, is of course extremely slim.

Let’s not forget that we are “moral animals”, bound by moral decisions. Sure moral decisions are restricted to the individual’s moral frame, but still they remain moral, and chances are that the might be the best given the circumstances.

Could AI decide whether a ship should run aground on a soft sandy beach, plagued with tourists, or on a rocky shore with the risk of sinking, if a problem arises? Are we willing to trade ” human error”, which could be corrected amidst the action, for a “computer glitch” that lacks any morals? The recent crash of an airplane equipped with state of the art technology was due to a software glitch, not to the so called “human error”.

Reducing costs and optimizing services is fine. But reducing costs by doing away completely with ships crew is playing poker. We might be bluffed by reality.

How about piracy? Pirates might also learn how to trick an autonomous ship into running aground to steal its cargo or just to cause harm.

I have to repeat that I am not an expert in anything else but in Piloting ships through the Panama Canal. In spite of that, and I might be wrong, the shipping industry will have to deal for many years more to come with not throwing away the coffee machine from the bridge of their ships. Mr. Pilot might just need another cup. Black and no sugar, please.

…………………….

As I finished this writing article the USNAVY is removing digital controls on many of its boats since they have found them to be one of the causes of many accidents and close calls that have taken place recently.

by Captain Ricardo Caballero “Themaritimepilot” – published on 15 June 2020
www.marine-pilots.com