MTCAS Maritime Traffic Alert and Collision Avoidance System

Motivation

The current maritime navigation systems’ technologies (like ARPA: Automatic Radar Plotting Aids) issue alarms such as Closest Point of Approach (CPA) and Time to CPA (TCPA) to warn the seafarers about critical situations or potential conflicts. However, this technology does not depend on analyzing historical tracking data, planned routes data, external information, knowledge-bases or environmental DB. Instead, it just depends on the current observed dynamics of ships (current position, speed and course) in order to predict CPA and TCPA. This can lead to wrong and useless alarms because current dynamics can change before the predicted CPA (e.g. in the next planned Way-Point). Thus, useless alarms (e.g. conflict on land or conflict between tankers and/or containers on shallow water) can be issued by traditional navigation systems technologies. However, the traditional technologies may issue important alarms, but they can be very late in some cases to have suitable reactions. MTCAS aims at exploiting all available information (historical tracking data, current observed data, planned routes data, Knowledge-bases, external information and environmental DB) in order to give early and more accurate predictions and suggestions to avoid potential collisions.

Fördergeber

Ziele

MTCAS project aims at developing an intelligent maritime collision avoidance system with three different levels of alarms with regards to the level of danger. This should be similar to the collision avoidance system available in aircrafts, where a “last line of defense” alarm can be issued in dangerous cases and the pilot must follow the suggestions accordingly in such cases.

Technologien

Technologies from Artificial Intelligence (AI), Prediction and Knowledge Representation domains are suggested to be used in MTCAS project. For example, regarding to prediction based on historical tracking data, polynomial regression, Bayesian algorithm based on a Particle Filter (PF), Adaptive neural-based fuzzy inference system (ANFIS) and Genetic Programing (GP) are candidate technologies. Rule-based knowledge base technologies are suitable to represent and query the International Collision Regulations COLLREGs, and local rules (e.g. rules in a specific port or harbor).

Personen

Projektleitung Intern

Projektleitung Extern

Sven Rohde (Raytheon Anschütz)

Wissenschaftliche Leitung

Publikationen
COLREGs-Coverage in Collision Avoidance Approaches: Review and Identification of Solutions

Mazen Salous, Axel Hahn, Christian Denker; 12th International Symposium on Integrated Ship’s Information Systems & Marine Traffic Engineering Conference; 08 / 2016

e-Navigation based cooperative collision avoidance at sea: The MTCAS approach

Christian Denker, Michael Baldauf, Sandro Fischer, Axel Hahn, Ralf Ziebold, Elke Gehrmann, Mika Semann; Proc. of ENC'16; 05 / 2016

MTCAS - An Assistance System for Maritime Collision Avoidance

Christian Denker, Axel Hahn; 12th International Symposium on Integrated Ship’s Information Systems & Marine Traffic Engineering Conference; 08 / 2016

Partner
Raytheon Anschütz GmbH
www.raytheon-anschuetz.com/
Signalis GmbH
www.signalis.com
Hochschule Wismar
www.hs-wismar.de
DLR Institut für Kommunikation und Navigation
www.dlr.de/kn/

Ansprechpartner

Laufzeit

Start: 01.01.2016
Ende: 31.12.2018

Fördermittelgeber

BMBF

FKZ: 03SX405D

Verwandte Projekte

COSINUS

Kooperative Schiffsführung für nautische Sicherheit