Safety for autonomous ships: lessons learned from other domains
Dr Stig Petersen Senior scientist SINTEF NORWAY
For autonomous ships, it is commonly stated that to be acceptable for commercial use, they must be at least as safe as conventional vessels in similar service. Safety can in this regard be defined as freedom from unacceptable risk of harm to humans, property or the environment. Since autonomous ships remove the human as a safety barrier, relevant safety regulations must be enforced for the control and navigation systems that will replace the onboard personnel. This presentation addresses how the maritime domain can learn from other modes of transport when it comes to safe implementation and testing of autonomous vehicles.
IConIC – the pre-emptive fault recognition system
Dr David Garrity Chief technology officer STS Defence UK
Globally there is a significant drive to introduce more automation and autonomy in marine and maritime, particularly through the exploitation of recent advancements in the application of artificial intelligence (AI) to solve challenging problems. Much of the focus has been on solving automation/autonomy challenges for navigation systems, and there has been limited thought about how to approach the engineering spaces, particularly the main propulsion and auxiliary power units. This presentation showcases an AI-enabled condition monitoring system, developed over the past five years for all scales of marine diesel units, and rotating plant machinery more generally.
Monitoring and control for vessels with reduced crew or no crew
Pia Meling Vice president Massterly (a Kongsberg Wilhelmsen JV) NORWAY
The two major Norwegian companies – Wilhelmsen and Kongsberg – have joined forces to lead the development of autonomous shipping globally. The main purpose of their Massterly joint venture is to develop environmentally friendly, safe and cost-efficient logistics, enabling a shift in transportation from congested roads to the sea. The first Shore Control Centre (SCC) is under construction, from which we will provide 24/7 manned monitoring and control for vessels with reduced crew or no crew. This presentation will provide an update on that progress. Massterly is currently in the process of obtaining approvals from the maritime authorities to operate fully autonomous vessels.
Required systems for proper look-out
Etsuro Shimizu Professor Tokyo University of Marine Science and Technology JAPAN
COLREGs rule 5 (Look-out) defines the look-out as follows: 'Every vessel shall at all times maintain a proper look-out by sight and hearing as well as by all available means appropriate in the prevailing circumstances and conditions so as to make a full appraisal of the situation and or the risk of collision'. This presentation discusses the required functions for the system that conducts the look-out.
One Sea to work with the whole maritime logistics chain
Päivi Haikkola Ecosystem lead DIMECC / One Sea FINLAND
Jukka Merenluoto Ecosystem lead DIMECC FINLAND
One Sea – Autonomous Maritime Ecosystem is a company alliance with the aim of enabling commercial autonomous maritime traffic by 2025. The Ecosystem engages in various activities to enable reaching its goal. In its third year the Ecosystem is concentrating more on the entire logistics chains, and the presentation will detail how this will be done. One Sea also participates in regulation and standardization work for MASS (Maritime Autonomous Surface Ships). The presentation will introduce the status of international regulation work and describe how One Sea aims to affect it. It will also discuss the way forward in standardization.
Thursday 27 June
Vision Technologies 14:15 - 15:30
Machine vision and sensor fusion as a step to Level 4 autonomy
Michael Gordon Johnson CEO Sea Machines Robotics USA
Sea Machines will present the results of its collaboration with AP Moller-Maersk in the world's first deployment of an AI-powered situational awareness vision system on an actively trading container ship. It will present the capability of using RGB and thermal camera vision in varying environmental conditions to provide real-time visual target and traffic detection, classification and tracking using robust artificial intelligence built with large data sets and cutting-edge neural networks. The camera vision-borne targets are then fused with target information from conventional vessel instruments to create a more effective domain awareness system for crews and autonomous control systems.
Detection of small cross-section objects with 3D lidar and polarimetric imaging
Dr Jordi Riu Gras CEO Beamagine SPAIN
Safe and reliable autonomous navigation needs to detect and track not only large objects like harbor structures and other vessels, but also small cross-section objects like buoys, rocks and even floating debris. This challenging capability is made possible by sensing in very high resolution and combining complementary imaging modes like 3D plus RGB or polarimetric data. The lidar provide point clouds with 0.05º angular resolution in real time, and the polarimetric camera is able to locate objects over the water with outstanding reliability. A patented hardware approach fuses both bits of information without parallax error at all distances up to hundreds of meters.
3D lidar for autonomous vessel navigation in close quarters
Dieter Gabriel Manager EMEA Velodyne Europe GmbH GERMANY
The presentation will provide an overview of how lidar technology is developing from the current state of technology into a mass-produced product suitable for marine applications and enabling vessels to go autonomous. It will include functional principles of lidar with respect to autonomous vessels, strengths and weaknesses of different environmental sensors, lidar technology evolution – from sensor in research to industrial mass product, and use cases for 3D multi-channel lidar sensors for autonomous vessels and marine applications.
Please note: this conference program may be subject to change