The Road Towards Autonomous Ship Handling with Tugs
Are autonomous tugs feasible? A good question, but let us first make clear what autonomous vessels are, because the term autonomous is not always used in a correct way.
The following definition is being used: An autonomous ship is a marine vessel with sensors, automated navigation, propulsion and auxiliary systems, with the necessary decision logic to follow mission plans, sense the environment, adjust mission execution according to the environment and potentially operate without human intervention (E-Navigation 09/01/19).
This means that autonomous vessels operate without human intervention. Examples are the 85-meter ferry Folgefonn which visited three ports autonomously, including docking and undocking, with no human intervention, and the Yara Birkeland which is a self-docking autonomous unmanned container ship of 79.5 meters, to be delivered in 2020, and will gradually move from manned operation to fully autonomous operating by 2022. Also a number of small workboats are already able to operate autonomously.
Then the question again. Are autonomous tugs feasible? Up until now tugs used for “autonomous” tests were all remote controlled. Remote control could be a first step towards possible further automation of tug operations. How will the future situation look like, for instance, around 2050? What are the practical issues to solve on the way towards autonomous tugs?
It is first important to understand that tug behavior during ship assistance differs essentially from a free sailing tug or ship, because when assisting a ship the tug operates in close vicinity to the ship and when the ship has speed the water flow between may strongly affect tug’s behavior and position. This is the main reason why a number of practical issues need to be solved.
These challenges will now briefly be addressed.
• Interaction effects between tug and ships: Interaction effects are complex and difficult to predict instantaneously. As said, they may strongly affect a tug’s behavior and position when operating in close vicinity to the ship to be assisted. Operating under influence of these effects can be very challenging.
• Ship speed: Speed is a critical factor, as it affects strongly all interaction effects working on the tug.
• Situational awareness of geometric environment: Next to executing the tug assist orders, a tug in assist mode still has to take care of the physical environment (not so much other traffic which is covered by the main vessel) and avoid contacting the port infrastructure (breakwaters and quays in particular).
These challenges can be handled by remote-controlled tugs if controlled by experienced tug masters having experienced the interaction effects frequently and being familiar with the port infrastructure. This means suitable tugs and an optimal working control center are needed. As the interaction effects increase by ship speed squared speed should be monitored carefully.
These aspects have been explained in detail in the book Tug Use in Port (see www.tugandosv.com).
For autonomous tugs it becomes even more complicated; tugs and control systems should be developed such that it can manage the problems itself. This will be a challenge, and an opportunity at the same time, to model these effects and cope with them even better than tug masters can do. This could be achieved with advanced machine learning using computational fluid dynamics-based models. Once modeled, these interactions could be anticipated more effectively and even be used to advantage of the operation. The tugs could then also be added to the maneuvering control capabilities of the to be assisted vessel and fully integrated as an add-on capability.
There are more hurdles to be overcome. Namely, it is crucial that tugs, whether remote controlled or autonomous, can connect to the ship to be assisted in order to be able to transfer the generated tug forces to the ship, unless when pushing and now towline is needed. There are at least three crucial aspects to this:
The procedure to make a connection.
At present the first connection between ship and tug is made by a heaving line thrown from the ship onto the tug to connect the towing line, for which the tug must come very close to the ship. If this takes place near the bow or stern of a ship having headway it can be risky for the tug and its crew because of the earlier mentioned interaction effects working on the tug, particularly when ship’s speed is high, e.g. more than six knots. Serious accidents have happened.
It is a system unsuitable for remote-controlled and autonomous tugs. Therefore, instead of this old-fashioned heaving line system, a safer system is needed that does not force tugs to come (too) close to a ship at speed and that is workable at speeds up to 10 knots, and in sea conditions and all weather conditions, while remote control of such a system is possible, and if possible, also autonomous working of the system.
The connection itself.
At present the connection itself is a towline which, on modern tugs, can be set at an appropriate length. A future tug-ship connection should, as with present towlines on modern tugs, also be flexible in any direction, and it should be possible to adjust the length of the towline if necessary, for instance to limit the negative effect of tug propeller wash impinging on the ship’s hull.
Various new systems for making a connection and/or for a connection itself are being tested of which drones as replacement for heaving lines might be a good solution. Such a system can also be remote controlled.
Future ships manned or unmanned.
If ships are manned, say up till 2050, it would make it easier to make a connection in the case of a remote-controlled connection system. Seeing the risks, the costs and the limitations of autonomous large vessels trading the oceans, it is expected that these future ships will sail with a (limited) crew, although the tendency towards an ever smaller crew can be expected, also seeing the increase of automatic mooring systems.
Net tug forces should be known.
A proper balance of forces should be possible for tugs operating autonomously. Therefore, net tug forces should be known. For instance, when the tug propeller wash of a pulling tug is impinging on a ship’s hull, it creates a counter effect resulting in a loss of the generated towing force on the ship. The loss, which is hardly to predict, can be more than 50 percent. It might even be such that the ship moves in the opposite direction to the pulling force.
The crucial question to be answered now is whether tugs used for ship handling will become remote controlled and/or autonomous before 2050?
The possibility is there to operate ship handling tugs remote-controlled and perhaps even autonomously when also the earlier mentioned practical issues are solved. The central question is then: Are remote-controlled tugs wanted and needed?
If such tugs are wanted depends to a large extent on country, port, towing company and also on the shipping companies, and, of course, on the costs involved. If remote controlled ship handling tugs are considered, it is good to analyze first all the ship handling tasks the local tugs have to carry out, for instance apart from the usually ship assist tasks, also ship assistance through locks and bridges, at dockyards, in wave conditions, handling oil rigs and large modules, fire fighting, emergencies, etc. If remote control is found still to be an option, then it should be borne in mind that each tug needs a controller with extensive experience as tug master. Therefore, it might then be best to just have a very limited number of remote-control tugs with the rest being normally manned.
At the same time, studies about autonomous tugs will continue for which more and complex challenges have to be addressed, as those mentioned above, than just the connection problems which are already a real challenge. It may take decades before unmanned autonomous tugs are operational in ports and at terminals.
For more detailed information about the subjects discussed, please see the link: http://www.swzonline.nl/news/10184/road-towards-autonomous-ship-handling-tugs
The opinions expressed herein are the author's and not necessarily those of The Maritime Executive.