The topic of evacuation analysis is becoming increasingly important in the maritime field, especially after the recent approval of relevant amendments to SOLAS. These amendments make evacuation analysis in early design stage mandatory not only for ro-ro passenger ships, as in the past, but also for other passenger ships, constructed on or after 1st January 2020, carrying more than 36 passengers. Tools used to perform evacuation simulations are generally run in a non-interactive batch mode. However, the introduction of the possibility for humans to interactively participate in a simulated evacuation process together with computer controlled agents in an immersive virtual environment, can open a series of interesting possibilities for design, research and development. Therefore, with particular reference to the maritime field, the research described in this dissertation is focused on the development and implementation of a mathematical model for simulating the dynamics of evacuation processes, which also allows real time human interaction through the use of virtual reality. The developed mathematical model, which is capable of naturally embedding human interaction, was verified and validated through a series of tests and through comparisons with other models and experimental data, as well as by referring to the relevant guidelines proposed by the International Maritime Organization (IMO). Particular attention was given to the calibration and validation of the counterflow model, developed during the research activity, and to the analysis of flow-density relation. The possibility of real time user participation, consisting in the user taking control over an agent inside the simulation, was introduced along with a vibrotactile haptic interface which was created to enhance the user perception of the surrounding virtual environment. The developed tool and user interfaces were adopted in an experiment where the subject was immersed in a virtual environment and interacted with simulated agents. The analysis of experiments provided results on the effects of the developed haptic interface on the subjects’ behaviour. Moreover, the obtained data allowed comparing the behaviour of subjects with that of simulated agents. The mathematical model was subsequently extended with the introduction of ship motion effects on agents behaviour, considering that, in the maritime field, the platform is usually moving. Fictitious forces, in the developed model, are directly applied to the agents and might therefore modify their trajectories. This represents an added value of the proposed model, because, usually, the effects of ship motions are embedded in simulation models only through a speed reduction. The model was used to assess ship motion effects in some IMO test cases. Finally, the tool was tested on a specifically developed case targeting the maritime field whose geometry was ideated as a simplification of the general plan of a real cruise vessel. The evacuation simulations were run firstly without ship motions, then with some representative situations combining heel, trim and periodic motions and, finally, with motions due to irregular waves. Ship motions, in this latter case, have been generated considering a notational cruise vessel whose dimensions were in line with the cruise vessel the test geometry was inspired to. A model introducing ship motion effects on the control of the avatar was finally developed, together with an approach to provide perception of ship motions through the developed vibrotactile interface. Models and results presented in this dissertation provide new insight to the field of ship evacuation analysis and to the application of virtual reality in this field.

Evacuation dynamics in the maritime field: modelling, simulation and real-time human participation

MONTECCHIARI, GABRIELE
2018

Abstract

The topic of evacuation analysis is becoming increasingly important in the maritime field, especially after the recent approval of relevant amendments to SOLAS. These amendments make evacuation analysis in early design stage mandatory not only for ro-ro passenger ships, as in the past, but also for other passenger ships, constructed on or after 1st January 2020, carrying more than 36 passengers. Tools used to perform evacuation simulations are generally run in a non-interactive batch mode. However, the introduction of the possibility for humans to interactively participate in a simulated evacuation process together with computer controlled agents in an immersive virtual environment, can open a series of interesting possibilities for design, research and development. Therefore, with particular reference to the maritime field, the research described in this dissertation is focused on the development and implementation of a mathematical model for simulating the dynamics of evacuation processes, which also allows real time human interaction through the use of virtual reality. The developed mathematical model, which is capable of naturally embedding human interaction, was verified and validated through a series of tests and through comparisons with other models and experimental data, as well as by referring to the relevant guidelines proposed by the International Maritime Organization (IMO). Particular attention was given to the calibration and validation of the counterflow model, developed during the research activity, and to the analysis of flow-density relation. The possibility of real time user participation, consisting in the user taking control over an agent inside the simulation, was introduced along with a vibrotactile haptic interface which was created to enhance the user perception of the surrounding virtual environment. The developed tool and user interfaces were adopted in an experiment where the subject was immersed in a virtual environment and interacted with simulated agents. The analysis of experiments provided results on the effects of the developed haptic interface on the subjects’ behaviour. Moreover, the obtained data allowed comparing the behaviour of subjects with that of simulated agents. The mathematical model was subsequently extended with the introduction of ship motion effects on agents behaviour, considering that, in the maritime field, the platform is usually moving. Fictitious forces, in the developed model, are directly applied to the agents and might therefore modify their trajectories. This represents an added value of the proposed model, because, usually, the effects of ship motions are embedded in simulation models only through a speed reduction. The model was used to assess ship motion effects in some IMO test cases. Finally, the tool was tested on a specifically developed case targeting the maritime field whose geometry was ideated as a simplification of the general plan of a real cruise vessel. The evacuation simulations were run firstly without ship motions, then with some representative situations combining heel, trim and periodic motions and, finally, with motions due to irregular waves. Ship motions, in this latter case, have been generated considering a notational cruise vessel whose dimensions were in line with the cruise vessel the test geometry was inspired to. A model introducing ship motion effects on the control of the avatar was finally developed, together with an approach to provide perception of ship motions through the developed vibrotactile interface. Models and results presented in this dissertation provide new insight to the field of ship evacuation analysis and to the application of virtual reality in this field.
16-mar-2018
Inglese
Ship safety; evacuation analysis; ship motion; real time user parti; virtual reality
GALLINA, PAOLO
BULIAN, GABRIELE
Università degli Studi di Trieste
File in questo prodotto:
File Dimensione Formato  
PhdThesisGabrieleMontecchiari.pdf

Open Access dal 16/03/2019

Dimensione 44.28 MB
Formato Adobe PDF
44.28 MB Adobe PDF Visualizza/Apri

I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.14242/177842
Il codice NBN di questa tesi è URN:NBN:IT:UNITS-177842