Analyzing Billions of AIS Records in a Multi-Purpose AIS Data Warehouse

Nielsen, Kasper Suamchiang Hvitfeldt ; Olsen, Marcus Egge ; Egede, Peter Kiib ; Øztürk, Alihan

4. term

Software, Master

2022

Dette speciale præsenterer et multifunktionelt data warehouse til at lagre og analysere meget store mængder historiske AIS-data for skibsfart. Vi kombinerer en udvidet datamodel på både logisk og fysisk niveau med processer for data­kvalitet, rensning, rekonstruktion af trajektorier og ETL, samt en repræsentation af verden som geografiske celler med aggregerede oplysninger. For at demonstrere anvendeligheden udvikler vi en fuldstack-applikation (API og frontend), der understøtter batymetrisk analyse ved at generere et Minimum Depth Chart (MDC) fra 1,8 milliarder AIS-poster indsamlet over fem måneder i 2021 i danske farvande. MDC’en sammenlignes med en dybdemodel fra Geodatastyrelsen (GST), og resultaterne viser, at vores model kan udpege områder af interesse, der kan kræve yderligere eller nye batymetriske opmålinger. Feedback fra GST understøtter, at AIS-data kan bruges til hurtige og omkostningseffektive analyser af forhold til søs. Systemet fungerer som et proof-of-concept og er bevidst fleksibelt, så det også kan understøtte andre use-cases, fx trafikheatmaps baseret på ture og celler.

This thesis presents a multi-purpose data warehouse for storing and analyzing very large volumes of historical AIS data from maritime traffic. We combine an extended logical and physical data model with procedures for data reliability assessment, cleaning, trajectory reconstruction, and ETL, and represent the world as geographical cells with aggregated information. To demonstrate utility, we develop a full-stack application (API and frontend) that supports bathymetric analysis by generating a Minimum Depth Chart (MDC) from 1.8 billion AIS records collected over five months in 2021 within Danish waters. The MDC is compared to a Depth Model provided by the Danish Geodata Agency (DGA), and the results indicate that our model can highlight areas of interest that may warrant additional or new bathymetric surveys. Feedback from the DGA supports the use of AIS data for fast and cost-effective analyses of real-world maritime questions. The system serves as a proof of concept and is intentionally flexible to support other use cases, such as traffic heatmaps based on journeys and cells.

[This summary has been generated with the help of AI directly from the project (PDF)]

Download
View record in AAU Student Projects