ben/nmis-kafka-ingestion
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
Page: Usage of Tresco Data with respect to machine learning
Pages
Usage of Tresco Data with respect to machine learning
1 Usage of Tresco Data with respect to machine learning
ben edited this page 2026-05-26 10:03:40 +02:00
Table of Contents

Problem

The mayority of the ships tracked inside the tresco files don't broadcast their ENI within 30 second via NMEA 2000. The old approach just dropped all of these vessels. But we might loose relevant data by just stripping away all ships that dont send their ENI in time.

Questions for Matthias

-> Can u imagine these ships being relevant in a predictive forecasting manner? -> Is it important for us that we can distingiush between every stranger ship that we recorded from time to time?

Recommendation: Store each ais entry that we have in the time-series table but only detailed information about vessels that we have an eni for?