TY  - GEN
AB  - Future waterborne transport operations in short-sea, sea-river, and inland waterways can be performed by autonomous vessels. The automation of maritime shipping directly emphasizes reducing crew numbers, minimizing operational costs, andmitigating human error during the operation. Recent researches have focused on understanding autonomous navigation while the reliability of unattended machinery plant has received very little attention. This paper aims at developing a method for predicting the performance of failure-sensitive components that may be left unattended in autonomous shipping. The presented methodology adopts Bayesian Inference as the basis of the artificial intelligence for predicting maintenance schedules including repair, inspection, and irregular checks of unattended systems. A Multinomial Process Tree (MPT) is used to model the failures within the system, identify faulty components, and to predict their failure times. A real case study from a short sea voyage is adopted to demonstrate the application of the presented methodology. The results of this research will assist decision and policy-makers to prevent costly failures in Maritime Autonomous Surface Ships (MASS) and extend the service life of autonomous systems before any human intervention.
AD  - Technical University of Delft, Netherlands;
AD  - University of Strathclyde, UK
AD  - Queensland University of Technology, Australia
AU  - Abaei, MM
AU  - BahooToroody, A
AU  - Arzaghi, E
DA  - 2020-10-05
ID  - 7750
JF  - Conference Proceedings of iSCSS
KW  - Autonomous Shipping
KW  - Reliability Engineering
KW  - Multinomial Process Tree
KW  - Bayesian Inference Unattended Machinery
L1  - https://library.imarest.org/record/7750/files/iSCSS_2020_Paper_23.pdf
L2  - https://library.imarest.org/record/7750/files/iSCSS_2020_Paper_23.pdf
L4  - https://library.imarest.org/record/7750/files/iSCSS_2020_Paper_23.pdf
LK  - https://library.imarest.org/record/7750/files/iSCSS_2020_Paper_23.pdf
N2  - Future waterborne transport operations in short-sea, sea-river, and inland waterways can be performed by autonomous vessels. The automation of maritime shipping directly emphasizes reducing crew numbers, minimizing operational costs, andmitigating human error during the operation. Recent researches have focused on understanding autonomous navigation while the reliability of unattended machinery plant has received very little attention. This paper aims at developing a method for predicting the performance of failure-sensitive components that may be left unattended in autonomous shipping. The presented methodology adopts Bayesian Inference as the basis of the artificial intelligence for predicting maintenance schedules including repair, inspection, and irregular checks of unattended systems. A Multinomial Process Tree (MPT) is used to model the failures within the system, identify faulty components, and to predict their failure times. A real case study from a short sea voyage is adopted to demonstrate the application of the presented methodology. The results of this research will assist decision and policy-makers to prevent costly failures in Maritime Autonomous Surface Ships (MASS) and extend the service life of autonomous systems before any human intervention.
PY  - 2020-10-05
T1  - Predicting Future of Unattended Machinery Plants: A Step Toward Reliable Autonomous Shipping
TI  - Predicting Future of Unattended Machinery Plants: A Step Toward Reliable Autonomous Shipping
UR  - https://library.imarest.org/record/7750/files/iSCSS_2020_Paper_23.pdf
VL  - iSCSS 2020
Y1  - 2020-10-05
ER  -