Nov 02, 2017 09:43 JST

Source: Fujitsu Ltd

Fujitsu Field Trials Ship Performance Estimation Technology with Mitsui O.S.K. Lines and Ube Shipping & Logistics
Successfully predicts ship performance at high accuracy; margin of error under 1.5%, according to trial with ship's operational data in actual sea conditions

TOKYO, Nov 02, 2017 - (JCN Newswire) - Fujitsu Laboratories Ltd. today announced that it has conducted a trial of a ship performance estimation technology(1) together with Mitsui O.S.K. Lines (MOL) and Ube Shipping & Logistics, Ltd. Developed by Fujitsu Laboratories and Tokyo University of Marine Science and Technology (TUMSAT), the technology can predict figures such as ship speed and fuel consumption. The results of the trial showed that the technology is capable of estimating ship performance with an estimated margin of error under 1.5%, both on the ocean and near the coast, and in a variety of weather and wave conditions. With this technology, it is possible for operators to accurately predict ship courses that will reduce fuel consumption, leading to lower fuel costs.

Fujitsu Limited will include this technology as an option in Fujitsu Mobility Solution SPATIOWL, a cloud service that utilizes location information, aiming to offer the service in 2018.

Details of this trial will be exhibited at the 2017 Fujitsu Forum Munich, to be held November 8-9.

Background

Figure 1: Illustration of weather routing (Light blue color indicates regions with moderate waves, orange denotes regions with somewhat high waves, and pink shows regions with high waves)
 

Figure 2: Example of a predicted ship speed against main engine rpm and wind speed applying high-dimensional statistical analysis technology (Light blue dots indicate actual values)
 

 

 

Figure 3: Example of trial results (ship speed)
 

Figure 4: Example of trial results (fuel consumption)
 


If maritime companies can get an accurate handle on ship performance, including changes to ship speed and fuel consumption due to weather and ocean conditions, they can plot optimal ship courses in response to such conditions along the route (this technique is called "weather routing"). For example, companies can know in advance if a slight deviation from the shortest route could actually improve fuel consumption as it avoids strong wind and waves (figure 1). In May 2016, Fujitsu Laboratories worked with TUMSAT to jointly develop technology that could estimate a ship's performance with high accuracy. Fujitsu Laboratories' proprietary high-dimensional statistical analysis technology utilizes Fujitsu Human Centric AI Zinrai, Fujitsu's AI technology, to produce highly accurate estimates.

http://www.acnnewswire.com/topimg/Low_FujitsuMitsui11217Fig1.jpg
Figure 1: Illustration of weather routing (Light blue color indicates regions with moderate waves, orange denotes regions with somewhat high waves, and pink shows regions with high waves)

Issues

Fujitsu Laboratories confirmed that this technology was highly accurate with estimating ship performance, through trials using operational data collected by TUMSAT's small training vessels in Japanese waters. As part of commercializing this technology, however, it was necessary to test estimation accuracy against operational data from actual commercial ships at sea, in more varied weather and ocean conditions.

Details of the Technology Trial

Now, with technical assistance from TUMSAT, Fujitsu Laboratories' high-dimensional statistical analysis, using operational data collected from commercial vessels in actual operation by MOL and Ube Shipping & Logistics, was the basis for prediction of speed and fuel consumption of the vessels, and testing performed for the margin of error compared to actual results (Figure 2).

The operational data used in this trial is as follows.

1. Mitsui O.S.K. Lines, Ltd.

Ship type: Oceangoing vessel
Data: 30 data items, including wind direction, wind speed, main engine rpm, and fuel consumption
Time period: October 12, 2015 through December 13, 2015

2. Ube Shipping & Logistics

Ship type: Domestic vessel
Data: 14 data items, including wind direction, wind speed, main engine rpm, and fuel consumption
Time period: July 29, 2016 through August 22, 2016

http://www.acnnewswire.com/topimg/Low_FujitsuMitsui11217Fig2.jpg
Figure 2: Example of a predicted ship speed against main engine rpm and wind speed applying high-dimensional statistical analysis technology (Light blue dots indicate actual values)

Methodology of the Technology Trial

A total of 90% of the sample data was selected at random from the total amount of operational data used in the trial as training data. The remaining sample data was used as test data to calculate the mean absolute percentage error (MAPE), an index of error calculated using the formula below. In this evaluation, T refers to the number of samples of the test data, and t shows the number of the given individual sample within T, xt gives the value of the input data used to make prediction number t in the sample, while f(xt) gives the predicted value for ship speed or fuel consumption based on xt, and yt gives the actual value for ship speed or fuel consumption for sample number t. This error evaluation was repeated ten times, and the average of those results was taken as the estimated error rate compared to actual results.

http://www.acnnewswire.com/topimg/Low_FujitsuMitsui11217Formula.jpg

When performing this error evaluation for ship speed, the model learned the relationship between ship speed data and non-ship speed data, and then the results of this learning and the non-ship speed data were used to predict ship speed. A similar error evaluation was performed for fuel consumption.

Results of the Technology Trial

The results of this trial confirm that this technology can predict ship performance with an estimated margin of error under 1.5% (table).

http://www.acnnewswire.com/topimg/Low_FujitsuMitsui11217Table.jpg

A portion of the trial results for ship speed and fuel consumption are shown in figures 3 and 4. Figures 3(a) and 4(a) are time-series graphs of actual values and predicted values, showing them as almost identical, confirming high prediction accuracy. In addition, figures 3(b) and 4(b) show the distribution of actual values and predicted values. As shown in the figures, most of the data points are clustered near the red line, confirming that this technology provides stable prediction accuracy.

http://www.acnnewswire.com/topimg/Low_FujitsuMitsui11217Fig3.jpg
Figure 3: Example of trial results (ship speed)

http://www.acnnewswire.com/topimg/Low_FujitsuMitsui11217Fig4.jpg
Figure 4: Example of trial results (fuel consumption)

By using this technology to accurately comprehend the ship's performance in actual sea conditions, users can improve weather routing precision and contribute to lower fuel consumption.

Future Plans

Fujitsu Limited will include this technology as an option in Fujitsu Mobility Solution SPATIOWL, a cloud service that utilizes location information, aiming to offer the service in 2018.

(1) Ship performance estimation technology
Fujitsu develops High-Accuracy Fuel Efficiency Estimates through a ship's operational data (press release, May 10, 2016)
(2) Accuracy of ship speed
Percentage of the difference between ship speed estimated using this technology and actual ship speed measured at one minute increments
(3) Accuracy of fuel consumption
Percentage of the difference between fuel consumption estimated using this technology and actual amount of fuel used during one journey

Contact:
Fujitsu Limited
Public and Investor Relations
Tel: +81-3-6252-2176
URL: www.fujitsu.com/global/news/contacts/
Source: Fujitsu Ltd
Sectors: Cloud & Enterprise, Marine & Offshore

Copyright ©2024 JCN Newswire. All rights reserved. A division of Japan Corporate News Network.

Related Press Release


Fujitsu Tech Leverages AI and Underwater Drone Data for 'Ocean Digital Twin'
March 26 2024 09:24 JST
 
Fujitsu Limited announces recruitment plans
March 19 2024 09:34 JST
 
Fujitsu and AWS will launch a new Modernization Acceleration Joint Initiative to enable customers across industries to drive digital transformation with fast and secure legacy modernization
March 18 2024 15:12 JST
 
Japanese joint research group win Prime Minister's Award with ultra high-performance computing platform using jointly developed 64-qubit quantum computer
March 15 2024 09:19 JST
 
Fujitsu and Tokai National Higher Education and Research System collaborate on AI-based space weather research
March 14 2024 09:21 JST
 
GMO Research Activity Support & Technology launches brain MRI analysis tool developed on world class supercomputer Fugaku with Fujitsu Computing as a Service (CaaS)
March 12 2024 09:20 JST
 
Fujitsu and Carnegie Mellon University develop AI-powered social digital twin technology with traffic data from Pittsburgh
March 07 2024 09:25 JST
 
Fujitsu and Celonis Expand Strategic Global Partnership
February 26 2024 14:40 JST
 
Fujitsu introduces "Uvance Wayfinders", expanded and strengthened consulting capabilities to deliver cross-industry business value
February 22 2024 11:05 JST
 
Fujitsu delivers new supercomputer system to Japan Meteorological Agency to improve prediction accuracy for typhoons and torrential rain
February 21 2024 15:38 JST
 
More Press release >>

Latest Press Release


More Latest Release >>