01/10/2025
10:37 UTC+1
Page size: 1.024 Mo
load time: 0.04Ms
ce site utilise des cookies_

Return to port prediction algorithm

maritime activity icon in green
activity:
maritime
expertise:
Computer science
Tow boat sailing in an industrial port at sunset with several loading cranes at the dock.
plus iconplus iconplus iconplus icon

Optimize logistics and reduce costs by predicting the return to port.

context
Strong competition from large European industrial fleets makes the optimization of operations crucial to the economic sustainability of small-scale fishing.

The Etaploise Maritime Cooperative wishes to optimize its logistics operations by accurately estimating the date and time ships return to port, to guarantee the availability of the logistics chain.
People in the cabin of a boat with navigation screens and a view of the water and the hills outside.
This example of a predictive algorithm makes it possible to predict the precise times the boats will return to port. Indeed, due to the absence of a network at sea, captains cannot announce their arrival. The prediction of the arrival time therefore allows trucks to receive the peach fruit at the right time, thus optimizing logistics operations.

From a certain size of vessel, it is mandatory to have an automatic identification system (AIS) to identify it and transmit crucial information such as its speed, GPS position, heading, etc. It is from this data that we can automatically detect the various phases of the journey (transit, fishing, return, etc.) and predict the time of arrival.
missions
Data retrieval and processing
minus iconplus icon
Analysis of historical and real-time data from sensors to understand operations and lay the basis for models.
Big data analysis
minus iconplus icon
Exploring and contextualizing large volumes of data to extract relevant operational trends.
Automatic detection of transactions
minus iconplus icon
Development of an algorithm capable of automatically identifying key operations and deducing performance indicators.
ETA prediction
minus iconplus icon
Design of an algorithm for predicting the Estimated Time of Arrival using operational and contextual data.
Visualization of results
minus iconplus icon
Creation of a web interface to monitor, explore and analyze all the data and models developed.

operational detection and return-to-port prediction

AI models and a dedicated interface to analyse operations in real time, refine predictions, and optimise planning.
Is this solution of interest to you?

benefits
This example of a prediction algorithm allows better organization of resources at the port and optimized logistics for processing fishery products. This project was co-developed in partnership with several stakeholders:
01

Fishing patterns

02

Captains

03

Logisticians

Footer Gradient

a single conversation can spark Innovation