Ships demonstrate AI-driven autonomous routeing
June 15, 2022 Maritime Safety News
Completed voyages by ships using artificial intelligence (AI) for navigation this year have demonstrated how far the shipping industry has come towards autonomous sailing, but also how much further it still must go.
Container ships, a gas carrier and an unmanned research vessel have successfully crossed oceans and seas using AI for routeing, collision avoidance and safety decisions. These proved technology for AI-advanced situational awareness and autonomous sailing works, but at least one of these demonstrations showed the need for crews on ships.
Mayflower autonomous vessel sailed unmanned from Plymouth, UK 27 April to North America, with detours on the way.
This autonomous trimaran docked in Halifax, Nova Scotia 5 June, instead of its initial voyage destination of Plymouth, Massachusetts, US on its second attempt, having failed due to a mechanical failure in Q2 2021.
The AI navigation system worked as expected, but Mayflower made a detour to Azores islands for emergency repairs on its generator.
When back on its transatlantic voyage, the 15-m vessel developed a problem with the charging circuit for the generator’s starter batteries. ProMare, which developed the autonomous vessel with IBM, switched to a back-up navigation computer 30 May and charted a course to Halifax, which was closer than any US destination.
AI and edge computer AI Captain guided Mayflower and made real-time navigation decisions. Six AI-powered cameras, more than 30 sensors and 15 edge devices provided input in AI Captain for analysis. It made split-second decisions, rerouteing itself around hazards or marine animals, without human interaction or intervention.
Also in Q2 2022, Avikus, a Hyundai group subsidiary, successfully completed a transoceanic voyage of 180,000-m3 LNG carrier Prism Courage using autonomous navigation technology. SK Shipping’s LNG carrier was equipped with Avikus’ HiNAS 2.0 autonomous navigation system.
Prism Courage departed Freeport on the US Gulf coast 1 May on a 33-day voyage to Boryeong LNG terminal, South Korea, via the Panama Canal.
The ship sailed half of the total distance, around 20,000 km, using the autonomous navigation technology. HiNAS 2.0 creates optimal routes and speeds based on Hyundai’s integrated Smartship solution. AI recognises the surrounding physical environment, weather and wave heights and nearby ships, then controls the ship’s steering in real-time.
During this ocean crossing, Prism Courage was operated autonomously on the optimal routes, increasing fuel efficiency by around 7% while reducing greenhouse gas emissions by about 5%.
Avikus says HiNAS 2.0 accurately recognised the locations of nearby ships during operation to avoid about 100 collisions.
This voyage was conducted with class societies Korean Register of Shipping and ABS montoring it in real-time monitoring to verify the technology performance and stability.
Avikus plans to commercialise HiNAS 2.0 this year after receiving certification from ABS for the results of this self-propelled ocean crossing.
In Japan, 749-gt container ship Suzaku sailed a 790-km autonomous voyage from Tokyo Bay, avoiding up to 500 ships in its 40-hour journey. This project involved Orca AI, the Designing the Future of Full Autonomous Ships (DFFAS) and The Nippon Foundation. DFFAS consortium comprises 30 Japanese companies, led by NYK group companies Japan Marine Science Inc and MTI Co.
During this trial, Suzaku automatically performed 107 collision avoidance manoeuvres on the outbound voyage, with the programme director at the consortium suggesting the system avoided 400-500 ships as Suzaku sailed towards the port of Tsu Matsusaka in Ise Bay.
Orca AI’s navigation system was set up on the vessel to act as a watchkeeper and provide real-time detection, tracking, classification and range estimation on 18 360° onboard cameras, running day and night.
Algorithms were trained on data collected over a year from Suzaku to identify targets in the complex Japanese shoreline environment.
Following this trial, Orca AI gained a contract to provide its AI-powered automated situational awareness platform to Maran Tankers Management (MTM) to improve navigation safety in congested waterways. MTM will also use Orca AI to maximise voyage safety and operational efficiency for tankers under its management.
Orca AI will provide seafarers with navigation information, enabling them to make data-driven decisions in congested waters.
This platform uses purpose-built machine learning and computer-vision algorithms to provide fleet managers and shipowners with information on fleet safety performance, enabling them to identify areas for further improvement.
“With Orca AI, MTM fleet crews now have additional highly advanced navigational equipment to use which allows data-driven, real-time decisions to be made,” says MTM managing director Mark Pearson.
“Our safety-first approach, together with our openness to technological innovation, drives us to be on a constant lookout for cutting-edge solutions to lower safety risks.”
In April, Chinese firm Bestway Marine & Energy Technology reported the successful maiden voyage of Zhi Fei, an autonomous, electric-powered container feeder ship. It made this maiden voyage from Qingdao Port, Shandong Province, 22 April. This 8,000-dwt demonstration vessel, with 300-TEU capacity, was developed through a broad collaboration of Chinese institutions to provide significant knowledge for the future autonomous navigation developments.
The 117-m long ship can sail at speeds of 12 knots and has a cruising range of 4,500 nautical miles. It integrates an intelligent navigation system jointly developed by Smart Navigation, CSSC 704 Institute, Water Transportation Research Institute, Dalian Maritime University and other institutions and has three modes: manual driving, remote control and autonomous navigation.
Bestway led the vessel design, working with Shanghai Jiahao Ship and Ocean Engineering Design and Research Institute (under Tianhai Defense) and Dalian Maritime University. Qingdao Shipyard built the vessel.
Zhi Fei was launched April 2021, commenced sea trials June 2021 and completed technical evaluations March 2022. It will operate between Qingdao Port in Shandong Province and Dongjiakou, managed by Chinese firm Navigation Brilliance.
Bestway says the ship includes multi-party intelligent co-ordination for navigation, scheduling, maritime supervision and ship support.
DB Schenker has revealed its plans to run an autonomous coastal container feeder for its customer Ekornes between Ikornnes quay and the Port of Ålesund, Norway.
It will work with partners Ekornes, Naval Dynamics, Kongsberg and Massterly to develop a 50-m unmanned and zero-emissions vessel to carry 300 dwt of cargo along a 43-km route at a speed of almost 8 knots. This will be supervised from Massterly’s remote operations centre.
