Global maritime trade carries over 80% of world cargo, yet vessel arrival time (VAT) prediction remains highly inaccurate. Hong Kong Port experiences average ETA-ATA deviations of 13.8 hours, causing massive congestion costs and supply chain disruptions.
Current methods rely exclusively on either static ETA reports or dynamic AIS data, missing the complete picture. This fragmented approach ignores how ships actually navigate—constantly responding to weather conditions, sea states, and their own physical capabilities.
This study develops a multi-source data fusion framework that integrates four key streams: ETA baselines, real-time AIS movements, marine weather data (wave height, wind speed, swell patterns), and vessel physical parameters (VPP). OpenFE automatic feature engineering handles complex data interactions, while six machine learning models (XGBoost, Random Forest, LightGBM, LSTM, Transformer, TabPFN) are systematically compared.
Testing on Hong Kong Port data shows TabPFN achieves optimal performance with 2.88–3.42 hour prediction errors, which means 43%–47% improvement over ETA baselines. Weather factors occupy 3 of the top 15 important features, contributing 20% of predictive power. Surprisingly, traditional machine learning consistently outperforms deep learning on this structured maritime data. These advances enable optimised berth allocation, reduced port congestion, and more reliable logistics planning, supporting the maritime industry’s digital transformation.

Tugboat operations manage the safe and efficient handling of vessel movements in and around ports. Accurate detection of these operations is necessary for the monitoring, planning, coordination and optimization of port-related activities. As maritime traffic continues to grow, there is a need for scalable and automated methods to detect and analyze these interactions based on available AIS data. The ability to systematically map tug activity across ports worldwide has commercial relevance. For maritime service providers, this facilitates large-scale analysis, operational planning, resource allocation and market development. In particular, identifying patterns of inefficient tug use offers actionable input for fleet optimization tools, such as those provided by KOTUG Optiport.


% Gap / Problem Statement
Existing detection methods based on AIS (Automatic Identification System) data rely on heuristic, rule-based logic and typically require predefined port zones. This limits their use in unstructured environments and prevents global deployment. Most AIS-based machine learning models analyze vessels in isolation and do not capture ship-to-ship interactions, which are central to identifying tugboat operations. This work addresses that gap by explicitly modeling the interaction between tug and assisted vessel as input, enabling the detection of relational behavior. In addition, it targets the specific trajectory segments where towing occurs, rather than labeling entire vessel tracks.


% Purpose / Objective
This thesis aims to develop a scalable, data-driven machine learning pipeline that can detect and classify tugboat operations worldwide by explicitly modeling ship-to-ship interactions using AIS data. The goal is to move beyond static, zone-based rules and build a method that uses dynamic behavioral data to identify interactions. The final objective is to implement this pipeline as a system that runs continuously and can be used in any maritime region across the world.

Container terminals are vital hubs in global trade, facilitating the seamless transfer of goods between maritime and inland transportation networks. Truck turnaround time serves as a critical performance metric for container terminals, provides direct feedback on port congestion and efficiency. This study focuses on developing a predictive framework for truck turnaround time (TTT) at the Port of Rotterdam by integrating multi-source data and employing advanced machine learning techniques.
Previous studies on truck turnaround time prediction have largely relied on limited datasets and methodologies, such as utilizing historical truck arrival flows or terminal operation logs and using statistical methods. This research employs diverse datasets, including Bluetooth detection records, container arrival information, and environmental condition. By combining these data sources, a harmonized dataset was constructed to represent the complexities of port operations. A stacked Long Short-Term Memory (LSTM) network was employed as the predictive model, utilizing its ability to capture temporal dependencies and nonlinear interactions between variables. This approach allows for more comprehensive and accurate TTT predictions compared to conventional methods.
To process the noisy and incomplete Bluetooth data, a robust trip identification pipeline was developed. The pipeline employed spatial clustering, temporal filtering, and dual verification to accurately identify container truck trips, achieving an accuracy exceeding 90%. Using this processed data, the stacked LSTM model demonstrated superior predictive performance, effectively capturing periodic trends and long-term dependencies. Benchmarking results showed that the stacked LSTM outperformed traditional methods, including Random Forest and XGBoost. Sensitivity analysis highlighted the critical role of truck arrival flows and wind conditions in determining truck turnaround time variability.
In summary, this study provides a novel and scalable framework for TTT prediction, integrating multi-source data and advanced modeling techniques to address key limitations of existing approaches. The findings offer actionable insights for optimizing terminal operations and reducing congestion. Future research could focus on expanding data sources, enhancing model interpretability, and validating the framework across diverse port environments to ensure broader applicability.

The transition to more sustainable operations is being widely adapted in order to reduce the green house gas emissions and meet future sustainability requirements. This transition most often utilizes electrification as a means to reduce emissions and utilize renewable energy sources. This transition comes with extra burden on container terminal authorities who have to manage their power demands and transmission and distribution system operators who have to keep up with providing this growing electricity demand.

This comes with extra costs as distribution system operators have to build and maintain a larger network and larger power capacities can not always be ensured for consumers such as container terminal authorities due to grid congestion. To achieve electrification for container terminals these distribution system operator costs as well as electricity costs and a congesting grid should be taken into account. To combat this, this thesis will analyze the electrification for a container terminal with a case study considering these factors.

However, scheduling power demands for container terminals is not trivial as they operate in a very dynamic and uncertain environment. This stochasticity is caused by uncertainty due to for example uncertain energy generation or uncertainty in operations, such as arrival time of ships. To ensure a container terminal has sufficient electric capacity and can manage its power demand for the day-ahead around this uncertain arrival time, a two-stage stochastic power optimization is modeled.

This optimization takes into account the flexible resources which a container terminal could benefit from, such as a battery energy storage system and flexible cooling of refrigerated containers. The charging decisions for the electric yard fleet as well as charging and discharging of battery energy storage system and cooling of reefers are scheduled for the next day. Power such as shore power and crane power for berthed ships which are loading or unloading are considered uncertain due to the uncertainty in arrival
and its deviation from the estimated time of arrival will be taken into account.

In this two-stage optimization where the aforementioned uncertain loads are second stage decisions, while decision such as when to charge batteries or cool refrigerated containers are made beforehand and therefore belonging to the first stage decisions. This stochastic two-stage optimization with uncertain ship arrival time is then solved with the progressive hedging algorithm, which decomposes the possible ship arrival scenarios in to individual solvable problems. These solutions are then pushed towards a common decision value through a penalty term.

With this model it is found that with the current electric contracted capacity, full electrification of the port equipment will not be a viable option. The necessary capacity is then optimized considering the flexible resources and electricity pricing. Dynamic electricity pricing will utilize a higher capacity to benefit from the lower electricity prices by charging and cooling at these times, despite the cost for a higher capacity. Despite these higher distribution costs, the total costs for electricity for a dynamic electricity price contract is significantly lower, minimally 23.65 % lower for the same configuration. A Time Constraint Transport Right is also analyzed, which could work for container terminals with many flexible loads, but this does not provide more incentive compared to a regular contracted capacity.

The current operational processes in an airport handling system (BHS) are not suitable for the implementation robots for the loading of baggage. This study aims to contribute to the implementation of new operational strategies, named batch-based pull approach, in a BHS to create a more automated and efficient operation. In this work a deep reinforcement learning (DRL) model is developed that can generate a baggage loading planning in real time for a baggage handling system in the dynamical operating environment to enhance robotic loading. The loading operation was formulated as a Markov decision process, and proximal policy optimization (PPO) algorithm was used to train the DRL agent. The DRL was compared with Vanderlande’s heuristic and tested on a case study of Brussels Airport. It automatically learned how to make baggage load planning decisions in simulations of a real-world BHS, generally loading more bags with a robot, but used more load units (LU), highlighting a trade-off between robot use efficiency and LU usage. This study demonstrates the potential of using deep-reinforcement learning for real-time loading planning in dynamic baggage handling systems with loading robots. However, more work is needed for consistent performance and real-world implementation. The results obtained are strongly related to the current model formulation, necessitating
additional research to gain more insight into the operating performance. This research serves as a proof-of-concept for future applications.

Efficient Port Operations are essential for minimiming delays and ensuring the safe
movement of vessels. Tugboats play a critical role in assisting ships during berthing,
unberthing, and navigating port waters. This thesis uses DRL4Route, a deep reinforcement learning (DRL) framework aimed at optimising tugboat routes and pick-up
locations. By using historical towage data and adapting to the dynamic conditions of
the Port of Rotterdam, DRL4Route provides real-time recommendations that streamline tugboat operations, reduce delays, and improve safety.
The core of the DRL4Route framework lies in its ability to continuously learn and adapt
to real-world conditions. Unlike traditional static models, DRL4Route leverages spatiotemporal data to predict optimal tugboat routes. This allows for real-time evaluations
that can help the predicted route match closely with the actual route which helps tugboats arrive at the right location at the right times. The framework’s focus on optimising
both pick-up and drop-off points helps port operators avoid inefficiencies that can arise
from poorly coordinated tugboat movements.
By improving the efficiency of tugboat operations, DRL4Route contributes to a more
sustainable and resilient port ecosystem. The system’s adaptability and potential for
real-time decision-making make it a strong candidate for future automation of tugboat
operations. This thesis highlights how advanced machine learning techniques can
enhance the performance of ports like Rotterdam, driving economic benefits and reducing the environmental impact of maritime operations.

Seaports are critical to the global supply chain, and as maritime trade grows with globalization, inefficiencies in port call processes and ship waiting times are increasing. This research develops a framework for Port Call Optimization (PCO) models to support collaboration and decision-making between stakeholders within the Nautical Chain (NC) and terminals. The Design Science Research (DSR) approach provides a systematic way to design a PCO tool for a port that is backed by a strong knowledge base. Integrating this approach with the co-design process involving user input ensures that the tool aligns with client needs and trends. This study uncovers a framework to assess the prerequisites for PCO tool implementation and develops steps to build a roadmap for PCO tools that incorporate the intricacies of different port governance structures. The findings from this research lays the groundwork for future research regarding the implementation of a PCO tool by bridging the gap between the fragmented research in this domain.

This paper investigates the optimisation of Automated Storage and Retrieval Systems (AS/RS) in warehousing by minimising performance losses during partial downtime. Given the increasing automation in logistics, AS/RS systems play a pivotal role, yet the operation of those systems during partial downtime remains a topic ignored in literature. This research fills this gap by exploring the effects of partial downtime in AS/RS through a reusable Discrete Event Simulation model which was developed in Python. This model incorporates the influence of both upstream and downstream systems, a characteristic notably absent from the limited number of publicly-available AS/RS models. Collaborating with Jumbo Supermarkets, the study utilises their highly automated distribution centre with an Order Consolidation Buffer housing 4 dual-crane AS/RS units as a case study. The study identifies operational policies to mitigate partial downtime effects, developed for scenarios with one or both cranes down within an AS/RS. Results suggest strategic workload distribution adjustments among AS/RS can significantly reduce performance degradation, particularly during high workload periods. After comparing both scenarios, it was concluded that for most scenarios, it is beneficial to keep operating the remaining crane when a crane breaks down, even though this slows down repairs. Overall, this research offers insights into parallel AS/RS dynamics under partial downtime and provides practical guidelines for effective operations.

Efficient production capacity planning is essential for industries facing uncertain demand, such as semiconductor manufacturing. This thesis explores machine learning models to improve demand forecasting and optimize production capacity, focusing on the ASML Field Service Hub. Models including XGBoost, Random Forest, and ARIMA are evaluated to predict demand across multiple product flows under uncertainty. These forecasts are integrated into an optimization model to minimize costs related to overcapacity, undercapacity, and labor fluctuations.
The proposed framework demonstrates substantial improvements over ASML’s current benchmarks, reducing forecast errors by up to 37.4% and simulated costs by 48.3%. Additionally, robustness testing was conducted using jitter tests, ensuring the model’s stability in the face of noisy data. The explainability and feature importance of the machine learning models were investigated using SHAP values. The study’s results highlight the model’s adaptability, offering a valuable tool for improving capacity planning in various industries dealing with demand volatility.

Due to the increasing trade of bulk cargo and the growing attention to efficiency in all steps of a supply chain that continuously becomes more complex, a thorough understanding of dry bulk terminal operations is crucial. While research on dry bulk terminal operations exists, it remains scarcely comprehensive. Existing studies primarily focus either on the storage space allocation problem, often in conjunction with berth allocation or on the simulation of the system, usually built for a specific case study. Furthermore, environmental factors have received limited attention in this context. Incorporating environmental considerations into this topic is crucial for sustainable and efficient terminal operations, especially when dealing with such a vastly spread sector as bulk cargo handling.
This work proposes a generic simulation model for grain terminals featuring a storage space allocation heuristic and the option to include cold ironing operations and unloading operations involving wind-assisted carriers. The generic character of the model is obtained by understanding what processes are common to different terminals and to what minimal level of detail they have to be modelled in order to obtain reliable simulation results. Moreover, the model gains further versatility due to its parametric and period-based character. These two features enable users to effortlessly conduct reliable simulations throughout all stages of design or revamping projects. This includes using information typically available at the start of a project as well as more detailed data that becomes available in later phases. The model includes the possibility to visualise and investigate the energy consumption of the system, crucial information to face present-day challenges such as the adaptation and enlargement of the electrical grid and the capacity estimation for the installation of new green energy production plants. The effectiveness of the model and its generic quality are validated with different study cases from real-world terminals. Additionally, in order to show the genericity of the model and its potential, this study features multiple experiments to investigate different aspects of terminal operational and energetic efficiency, involving the changes in operations caused by the introduction of shore power connections and wind-assisted carriers, a comparison of different unloading technologies, and a validation of the industrial common practices in the field.

Maritime shipping is essential for global trade, requiring the coordinated efforts of shipping lines, ports, and logistics providers. Early collaborative efforts focused on enhancing competitiveness, but the emphasis has shifted towards achieving environmental sustainability and resilience. This thesis provides collaborative approaches for resilient and decarbonized maritime and port operations, highlighting the importance of stable cost allocation methods to foster strong and lasting collaboration incentives.

This paper focuses on the effect of different storage policies on the performance of RoboticMobile Fulfilment Systems (RMFSs). The research is conducted under the instructions of the Technical University of Delft and CEVA Logistics the Hague. The aim of the research is to develop a decision support tool that can aid companies in the choice of the storage policy to use for their RMFS. RMFSs have multiple different levels of decision problems that need to be solved. The performance of a RMFS highly depends on the algorithms that are applied to solve these decision problems. This research focuses only on the storage policies of a RMFS. In this case storage policy refers to the decision in which pod items should be stored and where on the storage area the pod should be positioned. In order to gain a good understanding of the effect of such a policy on the overall performance of a RMFS, experiments should be performed. Physical experiments are however very hard and costly to perform. This research therefore makes use of a simulation study to test different storage policies in different scenarios. The simulation model used is an adaptation on an agent-based semi-open queuing network framework model by Merschformann et al. (2018a). In the experiments four different storage policies are investigated under three different storage layouts. The results of the simulations are analysed via the throughput, the pile-on and the distance travelled during the simulation. After this a score is given to
both the picking as well as the replenishment side of the system. It is important to investigate both sides of the system since the flaws on one side can negatively affect the other side. The scores of the replenishment and picking processes are then averaged to gain a final score which indicates the overall performance of the
policies. Finally a fifth policy has been developed where each pod can contain multiple different sized compartments. Unfortunately testing and verification of this policy was not possible in the given time frame. For this reason it has been left out of the experiments.

Electrification of numerous end-users is a worldwide trend to address climate change, according to the International Energy Agency. This trend has also reached container terminal operators. Currently most of the ship-to-shore cranes employed are electrified, leading to an increase in the required electrical power demand and to an increase in the volatility of the electrical power demand of container terminals. As a result, the contractual power demand charged by the grid operator, based on the maximum required power demand (peak power) at any moment in time, is upscaled, leading to additional costs for the container terminal operator. However, the highest required power demand values occur infrequently, leading to significant expenses for a resource that is rarely utilised. By implementing a peak shaving strategy, the peak power can be reduced, leading to a decrease in the contractual power demand related costs. Nevertheless, it is crucial to minimise the impact of the specific peak shaving strategy on the productivity of a container terminal to actually derive economic benefits from its implementation.

The aim of this study is to develop operational policies that effectively maintain productivity for a cluster of six ship-to-shore cranes under increasingly restrictive peak power limitations. A discrete event simulation approach was employed for evaluating the operational and economic impact. In total four policies were developed, two according to the `who fits is served' approach (policy 0 and policy 1) and two according to the `priority based' approach (policy 2 and policy 3). In the first approach the initiation of a movement only depends on the power availability, while for the second approach the initiation of a movement depends on the power availability and the urgency of the movement in terms of productivity. Moreover, for both approaches one policy allows only one kinematic profile (policy 0 and policy 2) and one policy allows varying kinematic profiles (policy 1 and policy 3). A metaheuristic was employed to find near-optimal adapted kinematic profiles.

The findings of this study suggest that the established `priority based' approach is more effective than the `who fits is served' approach in maintaining productivity under increasingly restrictive peak power limitations. When combined with the allowance of adapted kinematic profiles (policy 3), this strategy achieves the most cost savings. Policy 3, has been shown to reduce the contractual power demand related costs by 53\% compared to the baseline scenario, which is the greatest recorded reduction of all created policies without adversely affecting the ship-to-shore cranes' productivity.

Seaport operators are becoming more environmentally conscious and are looking to electrify their terminals to reduce their greenhouse gas emissions. This leads to higher energy-related costs and more congestion on the electricity grid. This thesis investigates the potential of demand response as a viable strategy to reduce energy-related costs. By modifying operational planning, energy consumption could be deferred from peak to off-peak hours, resulting in cost savings. Different potential ways within the terminal to provide demand response are identified. I propose a two-stage stochastic mixed-integer programming model to optimize operations planning, incorporating energy-related costs. Both energy demand and supply uncertainties are accounted for, exploring various scenarios for vessel arrival times and fluctuating electricity prices. The model is decomposed using a progressive hedging algorithm. Operational aspects considered in this model include vessel arrival scheduling, temperature control of refrigerated containers, allocation of handling capacity across quay cranes, yard cranes, and automated guided vehicles, as well as a charging schedule for the automated guided vehicles. A case study of the Altenwerder container terminal in Hamburg was conducted to test the model. Preliminary results suggest potential cost savings in the range of 12.0-13.2 % with a varying electricity prices based on wholesale market rates. Furthermore, it was found that stochastic modeling improved the solutions found of up to 20.6 % compared to a deterministic model. These findings underscore the substantial potential of demand response strategies in the context of container terminal operations

Fully autonomous bin handling systems are relatively new and examples of existing systems are hard to find. This makes designing such a system more difficult. The goal of this research is to provide insight in the effects of different design strategies of autonomous bin handling systems. This is done through a quantitative optimization model of an existing autonomous bin handling system. With this model, modifications to the system are made and evaluated. Optimal solutions for bill of material and floorspace, for different demand quantities are generated and compared.

The increase in online retail demand has stimulated automation in order picking systems, leading to new challenges and opportunities in task assignment and scheduling. In partially automated order picking systems, such challenges and opportunities exist regarding human factors implementation in the job-shop scheduling problem, an optimisation problem essential in operations. Workplace fatigue is a human factor often overlooked in scheduling research and application, despite hurting employees’ well-being and costing U.S. employers up to €127 billion annually. With the opportunities that automation offers, cobotic order picking systems could actively consider human fatigue development, mitigating its negative effects in operation.
This thesis investigates the possibility and potential benefits of fatigue consideration in the job-shop scheduling problem for a partially automated order picking system. We present a new bi-objective mixed integer nonlinear programming problem formulation to represent system constraints and a predictive fatigue model while considering worker fatigue and productivity during schedule optimisation. To put the results of simulated optimisation in perspective, we experimentally validate the fatigue model predictions and fatigue mitigation capabilities of the scheduling approach using heart rate measurements and qualitative fatigue ratings. These experiments occur with employees in a real-life partially automated order picking system.
Our mathematical model can find solutions that the conventional single-objective optimisation approach cannot, allowing fractional energy expenditure distribution improvements more than 4x larger than the decrease in productivity they require in 53% of the considered virtual cases. This is a promising result for fatigue mitigation in operations only by altering operational decision-making. However, the validation experiments show that our predictive fatigue model has an average RMSE of 2.20 kcal/min in estimating energy expenditure rates compared to heart rate measurements while also showing a low correlation. When assessing 10 minute intervals, a time span that fits a scheduling scope, the estimations improve slightly (avg. deviation of -1.85 kcal/min, avg. correlation of 0.17) but still underestimate the measured values. The experiments also show no significant differences in experienced fatigue between existing schedules and those with fatigue mitigation measures applied.
We conclude that the current scheduling formulation is not yet fit for application with a predictive fatigue model. However, real-life operations can benefit from energy expenditure estimation via heart rate measurements and a different approach for implementation is proposed. Research opportunities lie in further fatigue model development and validation, extension to indirect fatigue effects and other human factors, and further development of the mathematical formulation.

Inland waterway shipping, marked by its unpredictable and variable nature, plays a crucial role in transportation. This research's objective is to address these inconsistencies by constructing a robust scheduling model tailored to waterway systems' specific needs and challenges. The model is enhanced with predictive analytics and optimisation methods to ensure efficient and reliable operations. Within this framework, the research delves deep into four distinct optimisation problems: the Travelling Salesman Problem (TSP), the Job Shop Scheduling Problem (JSSP), the Resource-Constrained Project Scheduling Problem (RCPSP), and the Waterway Ship Scheduled Problem (WSSP). The underlying theme connecting these problems is the application of Graph Neural Networks (GNN) as a tool to model these complex systems.

The TSP is a cornerstone in combinatorial optimisation. Specifically, for this research, TSP serves as a means of understanding the challenges and intricacies of inland waterway networks. Employing the GNN model, the research provides insights into potential solutions for TSP within this context. When comparing various TSPLIB instances, the GNN model showcases its efficiency and potential for further refinement, especially in real-world routing and logistics.

Shifting the focus towards production planning, the JSSP emerges as a pivotal problem. It aims to optimise the order and timing of tasks for various ships, ensuring minimal usage of time and resources. By implementing the GNN architecture, the research offers a fresh perspective on JSSP. When applied to real-world scenarios, it is evident that the model can predict optimal scheduling sequences, matching the actual time frames and resource allocation required, thereby promising significant advancements in maritime trade efficiency.

Diving deeper into operations research, the RCPSP surfaces as a challenge that focuses on optimising project schedules, considering resource constraints and task precedents. The research introduces an approach to address this problem, especially concerning cargo operations within port networks. The research promises efficiency, reliability, and adaptability in Inland Waterway Transport (IWT) scheduling practices by integrating renewable resources and managing precedence relationships.

Lastly, the WSSP centres on managing ship movements within a defined time frame, optimising the sequence and timing of vessels to minimise delays and maximise the utilisation of waterway infrastructure and resources. Building upon the foundational work of previous research, this problem was translated and redefined in the context of the Resource-Constrained Project Scheduled Problem. Using this foundation, distinct RCPSP problems were formulated to reflect real-world scenarios, particularly emphasising the port of Duisburg. Drawing upon the results, the GNN model demonstrates high efficiency and accuracy in addressing the WSSP. While traditional tools like OR TOOLS provided optimal results, the GNN model closely mirrored these benchmarks, solidifying its position as a formidable solution for complex scheduling issues, especially given its rapid computation times.

In conclusion, this research presents a cohesive understanding of various optimisation problems within the realm of inland waterway shipping, all while harnessing the power of GNNs. Through systematic exploration and application, the research underscores the potential of GNNs to revolutionise how we approach and solve these challenges, promising a future of enhanced efficiency and reliability in waterway shipping operations.

Collaborative Port Call Optimization (PCO) is a key strategy to improve port efficiency. This study aims to bridge existing research gaps by providing a holistic understanding of collaborative PCO adoption, emphasizing the complexities of multi-organization collaboration and the influence of emerging technologies within the nautical chain and Port Governance structures. The TOEI framework, a novel conceptual model that extends the Technological Organizational-Environmental (TOE) framework, offers a valuable tool for port authorities and operators, providing specific insights and recommendations to enhance port call processes collaboratively. The study targets not only academics but also policymakers, port authorities, and nautical service providers. The findings discovered through the empirical analysis and the development of the TOE framework provide valuable insights for ports seeking to enhance their port calls collaboratively, laying the groundwork for future research, encouraging further exploration of specific contextual factors, in-depth case studies, and the development of tailored strategies for diverse ports globally.

The Automatic Identification System (AIS) is used in the maritime domain to improve sea traffic safety by requiring vessels to broadcast real-time information such as identity, speed, location, and course. As it allows global monitoring of almost any larger vessel and has the potential to considerably improve vessel traffic services and collision risk assessment, AIS has been used in an increasing number of applications. This emphasizes why the quality of data transmitted is critical. We are also becoming more aware of the possibilities of spoofing or fabrication of AIS data, which has a direct impact on the dependability of AIS data. This study looks into comparing video surveillance with Automatic Identification System (AIS) data to detect data overlap or spoofing in the maritime environment. An object detection model was proposed and trained to fill the research gap mentioned above. Data from video surveillance and AIS were collected. A comparison of the results of object detection and AIS data was performed using a Python script. The analysis assisted in identifying data variations and understanding the potential and constraints of the current Algorithm. Work for the future is suggested to tackle the current constraints.

The design of multi-robot systems has gained increasing attention in recent years. The field of cooperative Multi-Agent Robot Systems (MARS) has shown the potential to provide reliable and cost-effective solutions to a wide range of automated applications. Communication and coordination between autonomous agents require robust and intelligent control systems in order to achieve high-quality performance. This paper presents Collaborative Gym, an open-source, physics-based simulation framework for multi-robot interaction. This simulation environment differs from existing robotic simulation environments in that it is designed to model the interaction between multiple robots. Despite the presence of a large number of single robotic environments, multi-robotic simulation environments for reinforcement learning are rare. Collaborative Gym contains four simulated tasks in which different commercial robots work in collaboration: poking, lifting, balancing, and passing. For each of the four tasks, baseline policies are presented for various combinations of commercial robots which have been trained using reinforcement learning. The study demonstrated that Collaborative Gym is a promising open-source framework for the development of multi-robotic collaborative robotic tasks.