Visualising maritime vessel open data for better situational awareness in ice conditions. Authors: Jari Jussila, Timo Lehtonen, Jari Laitinen, Markus Makkonen & Lauri Frank. Academic Mindtrek Conference 2018, October 10.
Visualising maritime vessel open data for better situational awareness in ice conditions
1. www.hamk.fi
Visualising maritime vessel open
data for better situational
awareness in ice conditions
Jari Jussila † HAMK
Timo Lehtonen, Jari Laitinen SOLITA
Markus Makkonen, Lauri Frank JYU
2. www.hamk.fi
Motivation for studying maritime traffic in
ice conditions
1) In Northern Baltic Sea 21 ship-to-ship collisions occured during 2007-2013
3) A maritime vessel on a collision course in ice conditions has approximately only 3-4 minutes
time to react to the abnormal speed changes of the vessel in front in order to avoid a collision
2) Ship collisions are expensive and cause disturbances in the supply chain
3. www.hamk.fi
We claim that collisions can be predicted with open
data – and could be avoided by early warning
Source: MarineTraffic:
www.marinetraffic.com
Source: National Centers for Environmental
Information Weather and Climate Toolkit:
https://www.ncdc.noaa.gov/
4. www.hamk.fi
Proof-of-concept architecture for Ice Machine Learning
Marine open data from Finnish waterways: http://digitraffic.liikennevirasto.fi/en/marine-traffic/
Ice conditions open data from Baltic Sea: https://en.ilmatieteenlaitos.fi/ice-conditions
5. www.hamk.fi
Vessels getting stuck on ice
Vessel speed over ground (sog) 0-12
Ice thickness 0-1.4
A1 = vessel at port
A2 = speed falls to zero, possibly planned stop to wait for icebreaker assistance
A3 = small change on route
A4 = two sudden stops (getting stuck)
6. www.hamk.fi
A journey of three vessels ending up to
collision
A1 = Icebreaker Sisu (indicated by blue color)
A2 = Merchant vessel Aura, being towed by Sisu (indicated by orange color)
A3 = Merchant vessel BBC Virginia (indicated by green color)
Critical speed pattern
7. www.hamk.fi
Discussions and conclusions
• We introduced a solution that can provide real-time or near-real-time
descriptive, diagnostic and predictive analytics for improving maritime vessel
situational awareness in ice conditions
• We applied a triple open science principle: open data, open source
(https://github.com/timole/iceml/), and open access
• The introduced IceML proof of concept demonstrates that by enriching and
combining open data sources more valuable applications can be developed
• The introduced data science approach could also be used to improve the
quality of AIS data, e.g. user inputted vs data-driven estimated time of arrival
• Better situational awareness can lead to more optimal use of ice breakers or,
e.g., for companies to take proactive business actions in the supply chain to
mitigate the losses resulting from a ship arriving late