SAFETY4SEA Awards
2017
Learn More
SAFETY4SEA Awards
2017
Learn More

Analyzing Big Data –Managing data flows and using the information

BV

Antony Vourdachas, Senior Engineer, ABS, gave a presentation on “Analyzing Big Data –Managing data flows and using the information” during the 2017 SMART4SEA Conference & Awards. He argued that international regulations and operational costs are key drivers in today’s shipping industry in terms of vessel performance. Increasing regulatory compliance requirements add complexity to an already challenged market, while the volume of the data to satisfy such regulatory reporting is increasing. In parallel, the proliferation of low cost sensor technology and the interlinking of systems onboard a vessel have further increased the volume of data being collected and needing to be analyzed, he added. This leads to a challenge when handling, storing, analyzing and presenting such “big data”.  Mr. Vourdachas addressed some of these challenges and potential solutions to have in mind.

Regarding Big Data in shipping, there is a shift from Noon Data to Autolog meaning that 5 to 20 data points per 24h have become 10’s per 15 seconds. The increasing regulatory compliance framework is leading to increasing complexity, thus to more regulations and more data capturing. At the same time, there is an increase in the data amount; there is not only more data but higher frequency, leading to issues with handling, delays leading to data backing up etc. The traditional analytics, handling and storage is inadequate. In addition, trends can get masked by sheer volume of data. Transferring raw data from vessels is problematic and storing data in the physical sense is not practical. Therefore, now it is time for Big Data analytics, where all the data is being combined from the following data sources:  Performance Data; Consumption; Speed; Environmental Data; Machinery Data; Design Data and 3rd Party Data (AIS, Hindcast Weather, Machinery Data,…)

For the performance data, all events can be captured by Autolog, but a robust set-up is needed to alleviate the need for crew involvement in some of the capture, such as departure from port, ballast water exchange, bunkering, etc. Events can be captured for the whole voyage, from route planning to alongside but there is need to keep the crew in the loop by keeping them informed.

The data flows can consist of up to four source systems: Vessel Design Data, Autolog Data, Noon Report Data and 3rd Party Data (AIS, Weather, Design). These feed data to the Data Lake.

Data Consolidation happens in the central part:

  • Data Lake: Consolidation of data, data sorting, data validation. It is important to have rules in place to automatically match timestamps from different data sources (Autolog, 3rd Party, manual data)
  • Data Warehouse: it is the database which stores all the historical data, relevant data from sister vessels, etc. This communicates with the Data Lake and is a source for some of the data presentation.
  • Presentation layer: The data is fed to the Vessel Performance Management Software.

The analysis is standardized and automated, with the overview of the Vessel Performance team. Due to the size of the data involved, alarms and KPIs are used to simplify the analysis management task. The data is then presented in a standardized format with the possibility to be customized for each user (in terms of displays, KPIs, etc…). In addition the data should be presented in such a way to allow for ad-hoc analysis by the user.

Validation of Data consists of two stages:

1. Onboard: an important validation as it checks the data before going ashore for analysis. For the sensor performance, validation on spikes and outliers, drifting and malfunctions is needed. Any data found invalid, can be flagged and sorted into a location with other “problematic” data to be analysed and disregarded from the performance analysis. No data should be deleted. For the environmental sensors, there is need for validation using hindcast data, vessel motion sensors, etc… Validation rules should include the effects of the physical location of the sensor (i.e. obstacles which may block/accelerate the wind, etc…).

2. Data Lake: data can be validated with the use of a physical model (Calculated from vessel Design Data), a model based on data (Historical data. Requires historical data, or a period of data collection to generate the model) or a combination (refine the physical model using historical data).

For the vessel performance, the hull and propeller analysis is based on increase in non-dimensional friction coefficient which leads to increase in power and combining with SFOC leads to the increase in consumption. Main engine indices are based on M/E Manufacturer baselines, where available. There is need to identify state of ME Plant and effect on Fuel Oil Consumption and also to identify effect of corrective actions.

For the voyage performance you can capture all voyage events, SEEMP/EEOI/Emissions for planning, prediction and benchmarking. Include external data eg AIS, Hindcast/ Forecast and Design Data. For the environmental performance, reports can be generated for SEEMP, emissions reports, ECA areas & fuel switching overview, MRV & IMO data collection reporting. As all the data is collected, all environmental reporting can be generated.

For data presentation a standardized format may be used which simplifies the understanding of the large datasets and includes: Bar graphs, 3D plots, world maps – depending on the demands of the users. Filtering, overlays, Bin Ranges and Statistics can be used to generate simple reports from big data. Sorted/ filtered data is provided enable simple analysis and understanding of the data for users. Keep in mind the following:

  • Logging of data to a common database
  • Common interface for stakeholders
  • Categorize data
  • Link data to reporting
  • Integrate with analyses
  • Clear KPIs for decision support

In the future, we will see an integration of systems for Big Data in shipping which facilitates the structural monitoring, performance in a seaway and operational decision support. Also, we anticipate cargo tracking, remote operations, centralized knowledge and new vessel design. Cyber security is a new issue that industry has already taken into consideration and it is going to deal with in the years to come.

Above text is an edited article of Antony Vourdachas presentation during the 2017 SMART4SEA Conference & Awards

You may view his video presentation by clicking here

The views presented hereabove are only those of the author and not necessarily those of SAFETY4SEA and are for information sharing and discussion purposes only.

About Antony Vourdachas, Senior Engineer, ABS

Antony Vourdachas joined ABS in April 2014 as a Principal Vessel Performance Analyst in the Piraeus office.  Working for the Vessel Performance team in the Operational and Environmental Management department, he is part of the team responsible for providing the vessel performance analysis services to ship owners and operators and designing and updating the relevant software. In addition, he provides assistance to members of the Environmental team on MRV and ballast water issues. He started his career as a Design and Production Engineer at the Devonport Dockyard in Plymouth and moved on to become a Research Associate at the School of Marine Science and Technology at Newcastle University involved in a European Ballast Water project (MARTOB).  Prior to joining ABS, he worked for OSG in a number of technical positions in project engineering and planned maintenance. Antony holds a B.Eng. in Small Craft Engineering and an M.Res. in Marine Engineering from the School of Marine Science and Technology at Newcastle University, England.

 

Related Posts


Share on Facebook0Tweet about this on TwitterGoogle+0Share on LinkedIn7

Leave a Reply

Your email address will not be published. Required fields are marked *

*

Explore Our Group Sites:

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this. By continuing to use the site, you agree to the use of cookies. more information

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this. By continuing to use the site, you agree to the use of cookies.more information

Close