International shipping is involved in 90% of world trade. Just take a look at Shipmap.org to see the thousands of ships carrying millions of containers to get a sense of how truly enormous global shipping is on a daily basis. With significant opportunities to improve safety and efficiency in this huge market, a Rolls-Royce team based in Finland and Norway has been working to bring autonomy to shipping vessels using Intel® technologies.
Autonomous shipping presents many challenges. Disparate visual and positional sensor data must be detected, processed, and analyzed at the edge, all in real-time. With shipping, the data from these sensors are at greater scale than in other autonomous vehicle applications. For example, LIDAR sensors in shipping vessels must detect objects from several kilometers away, often in difficult weather conditions. Shipping vessels also need cameras with far-higher pixel count to see objects in, on, and over the water at distances of several kilometers, which means that large digital images must be processed for great detail. In addition, shipping vessels need to be outfitted with thermal, radar, and other sensors, all of which need to be fused together in order to read the environment. These sensors can produce enormous amounts of data – as much as 1TB of data per day, per vessel, about the same digital footprint as 250 high-definition feature-length movies.
To deal with this incredible amount of data, Rolls-Royce is incorporating Intel technology into their Ship Intelligence* platforms. Intel® Xeon® Scalable processors will be used on ships themselves and in data centers on land. Due to the need for real-time results, several server-class computers onboard ships will run machine learning and inference workloads locally.
Real-time inference on large images at scale is difficult. Deep neural networks must be employed to detect objects from far away and video encoding and decoding must occur. With the versatility for visual cloud, AI, analytics, and countless other applications, Intel Xeon Scalable processors give Rolls-Royce the flexibility and processing power needed for this type of extreme computing.
Shipping vessels are massive, allowing for the space and power to house modest on-board data centers. All of the data generated is stored on a “black box” (similar to an airplane’s flight data recorder), in case of a potential accident and for future training and analysis. Rolls-Royce currently uses Intel® 3D NAND SSDs for ship data storage and is evaluating Intel® Optane™ SSDs for future use. Additionally, Rolls-Royce is considering Intel® FPGAs to accelerate algorithm training using this data.
“This collaboration can help us to develop technology that supports ship owners in the automation of their navigation and operations, reducing the opportunity for human error and allowing crews to focus on more valuable tasks. Simply said, this project would not be possible without leading-edge technology now brought to the table by Intel. Together, we can blend the best of the best, to change the world of shipping.”
– Kevin Daffey, Director, Engineering & Technology and Ship Intelligence, Rolls-Royce
Rolls-Royce is planning for autonomous shipping fleets to be deployed by 2025. Though fully autonomous vessels are still several years in the future, technologies underpinning autonomous navigation already are showing value. Last year Rolls-Royce and global towage operator Svitzer successfully demonstrated the world’s first remotely operated commercial vessel in Copenhagen harbor. Additionally, situation monitoring technologies have been incorporated into Rolls-Royce’s Intelligent Awareness system, which was shown in a pilot program by Japanese shipping company Mitsui O.S.K. Lines, Ltd. to help the crew of a passenger ferry detect potential obstacles when navigating at night.
Autonomous navigation and intelligent awareness systems are coming to shipping vessels soon, and Intel technology will be at the forefront of workload consolidation, edge computing, communications, and storage. For more information about Rolls-Royce Ship Intelligence, visit them online. For more on Intel AI technologies, please visit https://ai.intel.com.