- Study Abroad Program
Download
Report
Transcript - Study Abroad Program
Shipping Routes Project
Scott Phan
An Nguyen
Presentation
27, June 2011
studyabroad.iit.demokritos.gr
Institute of Informatics and Telecommunications – NCSR “Demokritos”
Shipping in the Aegean Sea
Mediterranean Sea supports between 4-18% of the worlds
species
Aegean Sea is an area of the Mediterranean which carries
high biological importance, due to the relatively low coastal
development.
However, the preservation of this ecosystem is being left
largely to chance, with few protection measures in place. If
damage to the area increases or a major event the results
could be severe.
ML & DM Lab
27/06/2011
2
Shipping in the Aegean Sea
Hundreds of cargo, tanker and passenger ships pass through
the Aegean Sea every day. The potential impacts of
shipping, commercial and recreational, are vast.
Ships can affect marine biota in the following ways:
–
–
–
–
–
–
–
–
Underwater noise created by ships
Anchoring
Grounding
Direct collisions
Carrying invasive species
Operational oil discharges
Accidental oil discharges
Thermal Discharges
ML & DM Lab
27/06/2011
3
Project Motivation
The Aegean Sea lacks efficient mechanisms to manage,
monitor and regulate ship traffic conditions.
To decrease the chance of ecological disturbance events,
strict shipping lanes must be established and enforced.
ML & DM Lab
27/06/2011
4
Project Motivation
Over the last few months, the marine and GIS teams from
Archipelagos have been working on a major shipping project.
From November 12, 2009 to April 29, 2010, data on all of
the tankers and cargos that travelled between Samos,
Ikaria, Mykonos, Andros, and Nisos Evia were recorded.
The data was received from www.marinetraffic.com and
www.mariweb.gr, which track and record data from shipping
vessels.
ML & DM Lab
27/06/2011
5
Project Motivation
Accumulated Ship
Trajectories in the Aegean
Article:
“Update on Shipping Data
collection at Archipelagos”
-Chris Fletcher
Link
http://workjournal.archipelago.gr/?p=1348
ML & DM Lab
27/06/2011
6
Project Goals
Database
Design/Implementations
Import/Decode AIS data
GUI Design/Data
Visualization
Data Mining / Risk
Management
ML & DM Lab
27/06/2011
7
Project Goals: Details
1. Create a Graphical User Interface (GUI) [done]
2. Place ship markers at arbitrary pixel locations on the map [done]
3. Parse the AIS database and plot each ship at a given time interval on the
map [done]
ML & DM Lab
27/06/2011
8
Project Goals: Details
4. Create a function to assign a scalar “risk value” for each ship given
certain database attributes for each ship (cargo, size, proximity to
coast and other ships, flag, etc) [In Progress]
5. Create a visual identifier to show the risk of each ship (change
color, display numeric label, etc) [done]
ML & DM Lab
27/06/2011
9
Project Goals
7. Create a function to assess the risk of a particular map pixel
location based on the number and proximity of ships within a given
radius and proximity to land [In Progress]
8. Create a map view which shows risky areas in red and less risky
areas in green/blue. This “risk density map [In Progress]
ML & DM Lab
27/06/2011
10
Approach: Area of Interest
Spatial Bounds
Lat [35, 39]
Lon [21, 29]
Aegean Sea
Ignore all data that
falls outside the
boundaries
ML & DM Lab
27/06/2011
11
Approach: About the Input
Huge data collections
Good facilities to collect real time data
Accurate information to predict the trend of routes and risk
management
Used dataset provided by International Maritime Information
Systems (IMIS)
– Covers 2 days worth of AIS messages
Challenges with the IMIS dataset:
– Raw data - not decoded
– Not well managed
Redundant data (3x redundant)
– Database not as supportive for GIS development
ML & DM Lab
27/06/2011
12
Approach: About the Input
Automatic Identification System (AIS)
– is an automated tracking system used on ships and by Vessel Traffic
Services (VTS) for identifying and locating vessels by electronically
exchanging data with other nearby ships and VTS stations.
– AIS information supplements marine radar, which continues to be the
primary method of collision avoidance for water transport.
Types of Info. Encapsulated within the Messages
– Static [MMSI number, IMO number, callsign, ship name and
type, dimension]
– Dynamic [position, time, speed, heading, course over ground,
rate of turn, navigational status]
– Trajectory-based [destination, estimated time of arrival,
draught]
ML & DM Lab
27/06/2011
13
Approach: About the Input
Encoded Spatial Data within the Database
ML & DM Lab
27/06/2011
14
Approach: About the Input
Decoding the Encoded Spatial Data
Input :
0101000020E6100000000000604A653840000000E0FBB14240
Decoding
ML & DM Lab
27/06/2011
15
Approach: Software Architecture
Database
NASA 3D maps API
IMIS Database
Control
SQL Query
Raw Data
3D maps
Controller
Graphic User Interface
GUI
ML & DM Lab
27/06/2011
16
Approach: Graphical User Interface
ML & DM Lab
27/06/2011
17
Approach: Data Flow Architecture (Part I)
ML & DM Lab
27/06/2011
18
Approach: Data Flow Architecture
ML & DM Lab
27/06/2011
19
Approach: Output - Visualization
ML & DM Lab
27/06/2011
20
Approach: Output - Visualization
ML & DM Lab
27/06/2011
21
One more thing
27/06/2011
How we get data?
MySQL dump files
Extracted AIS Data
PostgreSQL/PostGIS
27/06/2011
How long does it take?
3hours/file
27/06/2011
SQL Importing Tool
27/06/2011
Demo
27/06/2011
Future Work
1. Prepare paper for publishing
2. Implement proximity function to find nearest
land from a given point in the sea
3. Design and implement the data mining API and UI
4. Research shipping risk assessment methods
5. Implement density plotting with respect to risk
assessment of a spatial area
ML & DM Lab
27/06/2011
27
Acknowledgements
Archipelagos - Institute of Marine Conservation
“KNOWLEDGE DISCOVERY FROM MARITIME MOVING OBJECTS APPLICATION TO AEGEAN SEA” – Cyril Ray, Naval Academy Research Lab,
December 2010
NCSR Demokritos
University of The Aegean
International Maritime Information Systems (IMIS)
CSE Dept. @ University of Texas at Arlington
ML & DM Lab
27/06/2011
28
Questions?
27/06/2011