Technology developed for bridge decision support will drive the transition to remotely controlled and then fully autonomous vessels
Technology for autonomous shipping is being developed for decision support on bridges and for fully unmanned operations, with computers bolstering situational awareness for navigators. Developments for short-route passenger vessels and inland tugs are focused on minimising crew or removing them altogether.
ABB Marine & Ports head of regulatory affairs Eero Lehtovaara expects developments for international shipping to incorporate more sensors and artificial intelligence (AI) to support navigation using camera feeds, light detection and ranging, radar and automatic identification system (AIS) information. These measure distances and detect obstacles through sensor fusion and mapping, assisting navigation teams.
“Bridge crew have several functions, including detection, identification, register and logging,” says Capt Lehtovaara. “We analyse decisions and then need to act – fast or slow – depending on the situation. But we will have machines in the mix in the future.”
These could perform collision avoidance, risk assessments and create action plans based on the situation.
“In most cases, human-in-the-loop will continue and technology on board will be for decision support,” says Capt Lehtovaara. “We will see changes in how people work on board.”
There are different levels of autonomy, with manual operation in the past and present and human navigation with decision support used currently. In the near-term, Capt Lehtovaara expects human-supervised automated functions to reduce the number of seafarers on ships. “We already have unattended machinery spaces, and this could be on the bridge in some conditions,” he says. “When operating in open seas, crew could work office hours and live in social norms.”
For this stage, supporting technology needs to be tested and matured and shipping needs confidence in its operation. In the mid-term, Capt Lehtovaara says ships could be periodically unmanned with human-supervised autonomy, which would involve remote control from shore and a back-up crew on board. Fully autonomous ships and operations are for the long term.
It will be a gradual development and technology trend. “There is no need to jump from being in the Stone Age to going to the moon,” says Capt Lehtovaara.
Part of the gradual transition will involve convincing regulators of the technology’s safety. “Autonomous and remote-controlled ships need to be as safe as conventional ships of the same type,” he continues. “Regulatory bodies want to see the safety and will need to be sure before technology can be allowed.”
Benefits of autonomy
For now, technology already developed will improve seafarers’ situational awareness and provide more motion control, vessel handling and machinery control. But more autonomy will benefit freight transportation and infrastructure. Remote operations will increase safety and sustainability, lower operating costs, and increase environmental protection and resistance to disruption.
However, there remains resistance to adopting such technologies due to concerns over safety, security, levels of investment and variable regulatory regimes.
“Some of the concerns about widespread use of autonomous transport methods, safety and security for instance, can be improved in certain circumstances through the technology,” says TT Club’s managing director of loss prevention Mike Yarwood. “An understanding of the implications of automation is vital for its responsible development.”
Coronavirus has acted as an accelerant in adopting autonomous technology, with a steep rise in using aerial autonomous vehicles for class surveys and infrastructure inspections. It has also renewed interest in remote hydrographic surveys and onboard engineering support. “But significant barriers still need to be overcome,” says Mr Yarwood.
One of the major challenges is the lack of uniform regulation across national governments and even within countries, which is a major block to autonomous vehicle and drone deployment. “Regulators have not kept pace with commercially driven technological advancement,” says Mr Yarwood.
Environmental hazards – bad weather, winds and high seas – affecting drone operation and autonomous ships are challenges technology can cope with. However, technology adoption needs to overcome the threat of cyber attacks on computer-controlled vehicles, as this is a strong disincentive to adoption.
“This particular threat is of critical concern across the supply chain in general,” says Mr Yarwood. “And autonomous transport with all its environmentally sound and economically tempting characteristics will continue to progress.”
Wärtsilä manager for strategic innovation Hussain Quraishi says the near-term benefits of autonomy at sea are increased safety and voyage optimisation. “These have already been realised,” he says. The medium-term benefits will be reduced crew on coastal cargo vessels, where seafarer expense forms a higher percentage of operating costs and “where enhanced situational awareness and precise manoeuvrability is at a premium” he explains.
Wärstilä’s autonomous technology in the marine sector is well advanced across smart sensors, optimised routeing and remote vessel control. “Our technology is demonstrable and has been proven to enhance safety and provide operational savings,” Mr Quraishi says.
Its technology will be used on the Mayflower autonomous ship project in September, when an unmanned and electric-powered ship will attempt a transatlantic voyage from the UK to the US. Mayflower will have Wärtsilä RS24 high-speed, high-resolution K-band radar (operating on 24 GHz) on board. This will work in tandem with Mayflower’s onboard cameras, ExactEarth AIS information and navigational systems as a core part of the AI Captain artificial intelligence control system.
AI Captain constantly evaluates Mayflower’s environment and long-term goals and modifies the ship’s course to avoid debris and storms which could threaten the ship at sea. IBM is providing the AI, cloud and edge computing, and power systems behind Mayflower’s AI Captain.
The RS24 has a five-times higher resolution than existing marine S- and X-band radars with spin cycles of 60 rpm. Wärtsilä said this will enable resolution separation of small craft in crowded waters, especially close to the vessel, and will enable Mayflower to navigate safely in complex situations.
This radar system is designed to identify potential hazards and uses data from the video recognition and AIS to produce a high-fidelity map of the operating area.
It was previously installed as one of the primary sensors on board Wärtsilä’s IntelliTug project, which underwent a trial in Singapore this year on board PSA Marine’s harbour tug PSA Polaris.
Reliable, cyber-secure communications will become increasingly vital for data transfer and connecting ships with ports. Ships will need to exchange more digital information with VTS centres, port and terminal operators and marine service providers in the future.
Connectivity requirements will change with the different levels of autonomy says
Inmarsat head of maritime digital Marco Cristoforo Camporeale. He says there are four degrees of maritime autonomous surface ships.
In the first degree, seafarers have complete control of ships, but with decision support from shore. The second degree involves shore centres remotely controlling ships, with crew remaining on board for maintenance and ready to take back control.
Mr Camporeale says the third degree involves ships being controlled from shore, with no crew on board and the fourth degree has fully autonomous ships with computers taking navigation decisions, but with an experienced officer monitoring operations from shore.
In the first-degree scenario, there are thousands of sensors on board capturing and then transferring data to shore and full crew on board, requiring communications.
“On Maersk Line’s container ships there are more than 2,800 sensors uploading 110 GB per month of data per vessel; Maersk downloads 30 TB [terrabytes] of data per month,” says Mr Camporeale. “They are saving US$20M in fuel costs over the entire fleet.”
Connectivity can help to reduce operational costs, create operational efficiencies, comply with regulations, create better value for customers and support crew welfare.
In second-degree autonomy, connectivity is required for remote monitoring and to control ships. “It will be important to be global with worldwide support, redundancy, availability, low latency and high capacity,” says Mr Camporeale. With crew on board to take manual control in emergencies, close to shore and in ports, communications will be required for operations and welfare requirements.
Ships with no seafarers on board would still need high levels of communications for large volumes of data transfers in both directions. In the third autonomous degree, ships are controlled from shore.
“When there is remote control of ships, loss of connectivity is not acceptable,” says Mr Camporeale. “Availability, low latency, high capacity and redundancy are important.” Connections would require bandwidth of 7-10 Mbps in both directions, depending on the use of edge processing on board.
When computers on ships are in control (Degree 4), they will still need connectivity to enable remote monitoring from shore. “Again, availability, latency and capacity are important,” says Mr Camporeale.
Inmarsat provides connectivity through Global Xpress satellites with Fleet Xpress, Fleet Data and Fleet Connect services.
Class evaluates NYK Line autonomous ship framework
ClassNK has developed new guidelines for automated and autonomous operation on ships and aims to establish evaluation methods for remote control and autonomous shipping.
According to ClassNK’s research institute manager Tomoaki Yamada its guidance, published in Q1 2020, covers design development, installation and operation of automated or remotely operated systems. It includes overall common basic requirements and procedures from the
viewpoint of safety verification at each of the lifecycle stages.
Also in Q1 2020, ClassNK granted approval in principle to NYK Line and Maritime Technology Institute for their joint project on the concept design of an autonomous ship framework (APExS)
This confirms the feasibility of the framework through safety evaluations that they are in line with the Guidelines for Automated/Autonomous Operation of Ships. ClassNK verified the validity and integrity of APExS through risk assessments, says Mr Yamada.
Its next contribution will be to confirm the validity and integrity of automated operating systems by using simulation methods.
ClassNK will also confirm the validity and integrity of remote operation systems. It needs to clarify the safety requirements for communications stability between the remote operations centre on shore and the vessel. “We aim to establish evaluation methods, tools and criteria for this, taking into account the knowledge gained in demonstration projects,” said Mr Yamada.
US$15M raised for autonomous naval vessel development
Sea Machines Robotics has secured US$15M of finance and formed a partnership with US military shipbuilding company Huntington Ingalls Industries.
This will accelerate the deployment of self-piloting technologies in naval vessels and commercial workboats. Sea Machines’ autonomous systems enable remote control of vessels from shore or a host vessel.
Its SM300 autonomous command and advanced perception technology has been deployed on a survey vessel operated by Deep BV for a remote survey of the Wadden Sea, north of the Netherlands.
Sea Machines Robotics has also deployed technology on a US-flag articulated tug-barge, oil-spill response craft, search-and-rescue, patrol and crew transfer vessels and a large Maersk cargo ship.