Over the past decade, autonomous surface vessels (ASVs) have increasingly operated in a range of challenging environments involving safety-critical scenarios. Their navigational capabilities rely on rich and reliable sensor data, enabling accurate localisation, situational awareness and environmental perception. This allows ASVs to perform motion planning, collision avoidance and navigational control tasks. To ensure maritime safety, faults affecting onboard navigational sensors must be diagnosed. This paper presents a model-based fault diagnosis scheme for ASVs affected by multiple sensor faults. Model-based methods utilise available sensors and dynamical models for residual generation. However, models describing the navigation may vary considerably for ASVs due to differences in vessel types, actuator configurations and sensor setups. To address this challenge, multiple residuals are synthesised using observer-based monitoring modules in the navigational sensors. Considering the impact of uncertainties, the residuals are designed to be bounded by adaptive thresholds proposed for each monitoring module. Fault isolation is then performed using a combinatorial decision logic, achieved by grouping the available sensors into multiple sensor sets and supported by model-based sensitivity analysis. Finally, the effectiveness of the proposed scheme is verified through simulation examples of two real-world vessels of different types with different sensor and actuator configurations, thereby illustrating its application. ... Read more

System identification of full-scale surface vessels must address significant uncertainties arising from model mismatch, sensor noise, and environmental disturbances. To provide safety, robustness, and constraint satisfaction guarantees, especially for autonomous navigation applications, it is essential to quantify the bounds of parametric model uncertainty. This paper proposes a set-membership identification method for estimating key parameters of a nonlinear vessel maneuvering model, including inertia and added-mass terms, other hydrodynamic derivatives in the Coriolis-centripetal, damping matrices, and actuation-related parameters. The method provides a bounded-error characterisation of uncertainties, offering a reliable framework for modelling the effects of measurement noise, wind, and waves. It involves computing a data-driven parameter set (DDPS) using input-output measurements and model assumptions, which is further used to compute a feasible parameter set (FPS). The parameter estimates are then obtained by iteratively solving a quadratic program over the FPS polytope. Validation of the method using experimental data from a full-scale catamaran demonstrates improved accuracy of up to 26.5% as compared to existing approaches, significantly faster computational times, and its capability to provide bounded parameter estimates. ... Read more

Inland waterways offer a cost-effective, energy-efficient and relatively safer mode of freight transportation, and autonomous navigation presents an attractive opportunity to fully revitalise their utilisation. To enable a safe, efficient and reliable inland waterway ecosystem, autonomous inland vessels must account for uncertainties arising from environmental disturbances and modelling errors. In comparison to open-sea navigation, inland navigation involves confined waterways, frequent interaction with infrastructure and other vessels, and operational constraints that require both high situational awareness and constraint satisfaction. Furthermore, abnormal operating conditions resulting from critical sensor faults and failures must be tolerated through graceful performance degradation or, in the worst case, a fallback operation. In light of these operational requirements, this thesis investigates how autonomous vessels, especially (but not limited to) inland waterway vessels, can maintain safe, high-performance motion control while monitoring sensor faults and hazardous situations that affect their navigation.

More specifically, this thesis contributes an integrated framework that includes (a) a robust system identification methodology to obtain vessel maneuvering models for state estimation and prediction, (b) a Nonlinear Model Predictive Control (NMPC)-based control system that computes the vessel's control actions while satisfying the physical and operational constraints of inland waterways, (c) a multiple sensor Fault Detection and Isolation (FDI) scheme that monitors consistency in measurements by employing analytical redundancy relations and (d) a risk mitigation method that provides a fallback control action under complex failures.

Robust system identification for marine surface vessels

Maneuvering models play a central role in model-based control and monitoring system design by providing accurate estimates of the vessel's states and their future predictions. Identifying the parameters of a full-scale vessel from experimental data is particularly challenging due to significant modelling and measurement uncertainties. The first contribution of this thesis is a set-membership method for identifying key parameters of a nonlinear 3-Degrees of Freedom (3-DOF) vessel model that supports robust prediction and control design through a bounded error characterisation of the uncertainties. The identification process involves computing two sets: a Data-driven Parameter Set (DDPS) and a Feasible Parameter Set (FPS), using the system dynamics, uncertainty bounds and input-output measurements. Then, by solving quadratic programs over the FPS, parameter estimates and their uncertainty bounds are obtained. Validation results from full-scale trials demonstrate improved prediction accuracy and reduced computational time. In addition, through sensitivity analysis, the parameters most crucial for identification performance are identified.

Path-following control of inland waterway vessels in confined waterways

Inland waterways are characterised by tight operational and environmental constraints, leading to explicit control design specifications. The model predictive control methodology is adopted, as it naturally integrates multi-variable dynamics, actuators, state, environmental constraints and objectives to optimise performance and control effort. An NMPC path-following control scheme is proposed for Inland Waterway Vessels (IWVs), with the prediction model tailored to the hydrodynamic phenomena in confined waterways, including bank and shallow-water effects. Many challenging scenarios are considered for validating the control scheme through simulations, such as turning at a steep river confluence, sailing a curved river and avoiding a static obstacle. The impact of reduced ship-bank distances, propulsion speeds and river cross-section shapes further provides insights into control performance and design choices. In addition, key performance metrics are proposed to evaluate the controller's performance and quantify path-following accuracy, robustness and safety.

Multiple sensor fault diagnosis of autonomous surface vessels

Autonomous vessels rely on multiple heterogeneous sensors for navigation, motion control and situational awareness. Sensor faults may propagate through measurements to interconnected systems on board, thereby impacting downstream decisions. This thesis proposes a multiple-sensor FDI scheme that exploits Analytical Redundancy Relations (ARRs) derived from the vessel's dynamical model and adaptive thresholds to diagnose sensor faults.

The design methodology adopted in the proposed scheme includes (a) the generation of fault detection residuals having structural sensitivity to one or more sensor faults and (b) the computation of adaptive thresholds used for residual bounding with robustness against environmental and modelling uncertainties. As a result, false alarms can be avoided in the fault detection process. In addition, a combinatorial fault decision logic is designed, enabling the scheme to not only detect fault occurrence but also to determine the compromised sensors. Combined, the structurally sensitive residuals and the decision logic facilitate the isolation of multiple sensor faults. The proposed fault diagnosis scheme is suitable for continuous monitoring of faults during vessel operation, while easily accommodating variations in the vessel's actuator or sensor configurations. Furthermore, by identifying weak fault sensitivity by evaluating residuals with respect to fault magnitudes, improved fault isolation decisions are obtained.

Collision and grounding risk mitigation of inland waterway vessels

Finally, the risk mitigation of autonomous vessels is explored by considering the underlying sub-problems of risk modelling and control. For risk modelling, a Bayesian Belief Network (BBN) is built from hazard analysis results, providing transition probabilities for sequential decision-making. Thereafter, a Partially Observable Markov Decision Process (POMDP) model is designed to represent the vessel's states and provide a suitable higher-level control strategy that ensures the vessel's safety by preventing hazardous situations, such as grounding and collisions. The method is verified through an inland waterway navigation case study, which demonstrates SCS selection reliably during a complex failure scenario.

... Read more

Autonomous inland shipping offers a safer and more efficient form of transportation over water with the potential to reduce maritime carbon emissions. However, the operation of autonomous vessels presents unique challenges due to complex dynamics, varying traffic conditions, and environmental disturbances. To ensure the safe navigation of these vessels in confined inland waterways, it is crucial to address manoeuvring prediction and motion control challenges. Research focusing on these challenges disregards or only partially incorporates inland waterway characteristics related to the vessel and its surroundings. This study provides a comprehensive analysis of these key factors. By modelling the vessel using a modified Manoeuvring Modelling Group (MMG) model specifically tailored for confined waterways, hydrodynamic effects due to shallow water, channel banks, and current are accounted for. A nonlinear model predictive controller (NMPC) is employed for the vessel path following control under various scenarios, including straight channels, confluences, and river bends. It is observed that the hydrodynamic effects from the channel banks significantly impact vessel steering. Compared to conventional proportional-integral-derivative (PID) controllers, NMPC effectively reduces course deviations and cross-track errors under varying water depth and ship-to-bank distance conditions, while also requiring fewer rudder deflections. Furthermore, key performance metrics related to the control of inland waterway vessels are proposed to evaluate the controller's performance further. The NMPC control law demonstrates its effectiveness in capturing the hydrodynamic effects and improving navigation safety in confined waterways. ... Read more

Autonomous surface vessels (ASVs) increasingly gain appeal in the maritime industry for their high efficiency and improved navigational capabilities. However, risks originating from various internal and external factors such as faults, traffic, harsh weather conditions, etc., can affect their guidance and control capabilities and impact nominal vessel operations. The existing risk mitigation methods mainly focus on the vessel's guidance system and do not consider unsafe actions due to the control system. In this paper, we propose a new method based on a partially observable Markov decision process (POMDP) model for the online risk mitigation of autonomous inland vessels. The POMDP model-based method utilizes information about situational awareness to assist the vessel's planning and control system in real-time decision-making during hazardous situations, thereby ensuring that the vessel remains in a minimum-risk condition. Based on the identified risk-influencing factors (RIFs), the transition probabilities are updated by a Bayesian belief network (BBN). A case study of an autonomous inland vessel navigating in a confined waterway is presented to demonstrate the capability of the proposed method. ... Read more

Autonomous surface vessels (ASVs) have started to operate in many safety-critical scenarios where rich sensor information is required for situational awareness, environmental perception, motion planning, collision avoidance and navigational control. A timely diagnosis of faulty onboard sensors is therefore essential for ensuring maritime safety and reliability. In this paper, a model-based fault diagnosis scheme is presented for ASVs affected by multiple sensor faults. Various monitoring modules comprising nonlinear observers are employed for the detection of faults occurring in the vessel’s navigational sensors. Further, multiple fault isolation is performed based on a combinatorial decision logic, achieved by grouping the available sensors into multiple sensor sets. The efficacy of the proposed scheme is illustrated through a simulation example of a vessel trajectory tracking scenario. It demonstrates the scheme’s ability to effectively isolate multiple fault combinations impacting the sensors considered. ... Read more

Autonomous inland shipping has great potential to enable intelligent and sustainable freight transport. At the same time, with the increasing traffic on confined waterways, ensuring safe operations of these autonomous inland vessels within limited operational spaces becomes imperative. This will require considering hydrodynamic effects during control design stages. This study presents a comprehensive analysis of an autonomous inland vessel’s manoeuvrability and controller design. The ship’s motions are modelled using an enhanced Manoeuvring Modelling Group (MMG) model to account for the hydrodynamic effects of inland waterways, including water depths, river currents, and bank effects. A verification study is conducted utilising a pusher-barge prototype model in shallow water to verify the model’s accuracy. Through the implementation of a controller, a course-keeping study is conducted to assess the vessel’s steering performance across various inland waterway scenarios, including sailing along river bends and waterway intersections. The results show that the manoeuvring model can generate fast and accurate vessel trajectory predictions. It is found that the proposed control technique proves effective in mitigating the confinement effects and countering disturbances caused by river currents, thereby ensuring efficient course-keeping suitable for the considered type of autonomous vessels on inland waterways. ... Read more