This dissertation addresses consumer decisions in last-mile logistics, with a particular focus on the emerging concept of crowdshipping. By conceptualising consumers as both users and service providers of logistics services referred to as “prosumers”, this research examines how their dual role influences the efficiency and sustainability of last-mile delivery systems. Through a combination of defining a conceptual framework, empirical analyses, and a simulation study, this work investigates key aspects of crowdshipping and the role of prosumers in supply and demand perspectives.

The dissertation is structured around a main research question: How does prosumer decision-making within crowdshipping impact the performance and sustainability of last-mile delivery systems? To answer this overarching question, the research investigates several key aspects of consumer behaviour, specifically focusing on how consumers make decisions within hyperconnected delivery networks and their evolving role as active participants in crowdshipping…

The decarbonization of heavy-duty road freight transport requires overcoming the dual challenge of limited driving range and insufficient charging infrastructure for battery-electric trucks (BETs). This thesis develops a nationwide bi-level optimization framework that integrates the deployment of static charging stations (SCS), electrified road systems (ERS), and truck-level battery assignment, leveraging large-scale tour-based freight data from the Netherlands. By operating explicitly at the tour level rather than the trip level, the framework captures cumulative energy feasibility across multiple linked trips, thereby providing a more realistic representation of freight operations. The model is solved using a Genetic Algorithm (GA), capable of evaluating more than 1.5 million tours and 3.5 million trips, and validated against exact MILP solutions on small instances. The results show that the optimization raises feasibility from 58% to 89.9%, reducing infeasible tours to 10.1%, while overall fitness improves by 34.7%. ERS emerges as the backbone of electrification, with 12,792 km deployed (€25.7 billion, 65.6% of CAPEX), while 251 SCS facilities (€50.2 million, <1% of CAPEX) provide low-cost redundancy at regional hubs. Battery allocation is highly heterogeneous: 25% of trucks operate on 90 kWh, 40% on 600 kWh, and the remainder on intermediate sizes, yielding an average of 357 kWh—closely aligned with ElaadNL’s benchmark of 289.5 kWh/day. This heterogeneity reduces battery CAPEX by approximately 19% compared to a uniform-capacity baseline. Operating expenditures (OPEX) are dominated by ERS charging, while penalty costs for infeasible tours remain substantial, averaging €362 per unserved tour. These findings demonstrate that nationwide electrification is feasible under a layered strategy: ERS as the long-haul backbone, SCS as regional redundancy, and heterogeneous batteries as cost optimizers. The study advances the literature by moving from trip-level to tour-level modelling at unprecedented scale, explicitly quantifying infeasibility, and decomposing system costs into CAPEX, OPEX, and penalties. The results provide actionable insights for policymakers and industry stakeholders seeking cost-effective and operationally viable pathways for freight decarbonization in the Netherlands and beyond.

Recent legislative developments, most notably the new Packaging and Packaging Waste Regulation (PPWR), state that by 2030 a minimum of 10% of beverages must be sold in reusable packaging formats, increasing to 40% by 2040. This makes a transition unavoidable, and requiring HEINEKEN to further increase the share of their volume sold in reusable packaging, to provide for the final distributors. Reusable glass bottles (RGB) are seen as the most promising material for this transition. Through a Design Science Research approach, this study developed two complementary artifacts. First, a conceptual model mapping eight strategic decision dimensions and their parameters, structuring the full configuration space for RGB supply chains. Second, a financial model translating this framework into a practical evaluation tool to evaluate configurations on performance and trade-offs financially, environmentally and operationally, and that quantifies outcomes such as cost per hectoliter, Net Present Value (NPV), Discounted Payback Period (DPP), profit per hectoliter, and CO2 emissions. The demonstration and evaluation confirmed that the models behave logically and align with theoretical expectations. Under benchmark conditions and at 500.000HL, the In-house configuration came out as most promising, with Hybrid close behind. Outsourcing can be the more attractive option at lower volumes, while at higher volumes the advantage shifts to In-house and Hybrid. Pooling consistently performed weakest, but could at lower volumes and lower fees be viable and benefits from the lowest operational effort. The impact analysis identified production strategy, crate strategy, and sorting \& washing as the most critical levers influencing performance and sub-configuration tests also showed that choices inside a strategy can have impact. It can be concluded that there is no single supply chain design that would always operate best. Outcomes depend strongly on volume and market-specific conditions. This research provides a structured configuration framework and evaluation model to support the transition towards reusable glass bottle supply chains.

This thesis examines the feasibility and innovation potential of battery swapping for electric vehicles in the Czech Republic, focusing on micromobility and urban logistics. Using the Technological Innovation Systems (TIS) framework, seven mobility segments were evaluated against key innovation functions. Findings show micromobility and parcel delivery as the most viable near-term candidates, with micromobility particularly suited due to low costs, high usage frequency, and alignment with Sustainable Urban Mobility Plans. A Business Model Innovation (BMI) framework is developed for a modular, multi-user swapping system integrating e-scooters, delivery services, and logistics operators, financed through hybrid public–private mechanisms. Expert validation highlights both opportunities and conditions for scalability, including standardization, operator collaboration, and policy support. The study concludes that battery swapping, framed as shared infrastructure, can deliver operational, environmental, and societal co-benefits, offering a systemic pathway toward sustainable urban mobility in Czech cities.

Misalignment between truck arrivals and terminal capacity is an increasingly critical bottleneck in major seaports within the global containerised supply chain. This misalignment is largely due to unpredictable delays caused by both road congestion and terminal operation disruptions. Traditionally, Truck Appointment Systems (TAS) aimed to coordinate pick-up and drop-off activities with the use of fixed time slot reservations. However, this approach lacks the flexibility to adapt to real-time disruptions, particularly in a siloed multi facility setting.
This study proposes an integrated and dynamic multi-facility coordination framework that applies real-time Estimated Time of Arrival (ETA) data into appointment rescheduling and reassignment decisions. Through a rolling horizon approach, schedules are continuously updated to reflect the most recent system state. The scenario analysis has demonstrated that the integration of real-time ETA data can reduce total waiting times by up to 96%, with the enablement of container reassignment resulting in the most efficient outcomes. Through multi-objective optimisation, the trade-off between minimising waiting times and limiting reschedules has been assessed. Sensitivity analyses further illustrated how congestion severity, timeslot duration, and fleet sizes influence overall system performance. The results indicate that collaborative, data-driven coordination mechanisms can significantly improve port logistics efficiency, reduce port congestion, and enhance service reliability. These findings provide a foundation for the development of dynamic decision-making models that balance operational stability for terminals with minimal delays for carriers.

Heavy-duty trucks contribute significantly to greenhouse gas emissions in the transport sector,
despite making up only a small portion of the total vehicle fleet. As the EU targets a 90% reduction
in truck emissions by 2040, Electric Road Systems (ERS) have emerged as a potential solution by
enabling trucks to charge while driving, removing the need for large, heavy batteries. However, little
is known about the temporal and spatial energy demand of such an infrastructure and the impact
it would have on the electricity grid—particularly in the Netherlands. This study investigates the
expected temporal distribution of energy demand resulting from an ERS corridor between Rotterdam and Venlo. A quantitative approach is used to model freight flow, ERS adoption potential, and estimate energy consumption. BasGoed origin-destination data is used to map freight volumes,
and a set of spatial criteria determines which trips are eligible for ERS usage. By combining the data and demand characteristics, a temporal and spatial demand pattern can be identified, that can be used by the industry and policy makers to understand ERS and what is needed to support future implementation.

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.

Private rail sidings form the physical access gates for companies to the rail network. Without these rail freight, transport is not possible and these private sidings can be seen as an indicator of the competitiveness of rail transport. However, the number of private sidings has dropped significantly over the last decades, which limits the rail freight sector and the resilience of the rail network. A resilient rail network is important to reach societal goals such as emission reduction and less road nuisance. This research focuses on this problem by proposing concrete strategies to help the companies using rail transport and allow them to use their private sidings. Through the help of literature, data and stakeholder interviews, an Osterwalder business model canvas, SWOT analysis and TOWS matrix are constructed to identify the problems and propose strategies. Causal diagrams show the coherence of the sector and how problems, strategies and private sidings influence each other. The first result is a new way of categorising companies and their respective private sidings. This reveals the underlying problems and strengths per type of private siding. The most important problems are rising costs, unreliability of the rail network and limited usability of the private siding due to regulations and non-electrified infrastructure. To combat this, six strategies are proposed with the introduction of a second infrastructure manager to combat costs and regulations and specific infrastructure investments as the most promising strategies. Effective policies require tailored solutions based on the type of company and private siding. This can help the Dutch rail sector and increase the resilience of the transport network.

Freight transport is vital for urban economic growth, yet its increasing demand exacerbates congestion and emissions. The integration of freight delivery within public transport networks presents a sustainable alternative, leveraging underutilised transit capacity. While prior studies focus on operational aspects, limited research addresses strategic hub locations within public transit networks. This study introduces a novel framework for selecting public transport stations as freight hubs using a Multi-Criteria Decision Analysis (MCDA) approach with the Best-Worst Method (BWM). The framework evaluates key factors, such as accessibility, capacity, and costs, to determine optimal hub locations. A simulation is subsequently conducted to assess real-world viability, demonstrating the trade-offs between network reach, handling efficiency, and financial constraints. The results indicate that strategic hub placement significantly impacts network scalability and efficiency. Future research can expand
upon this framework to refine urban freight delivery models tailored to specific implementation areas.

This thesis discusses the economic developments around 21 road and rail corridor projects in Africa. These projects were funded by development banks or by foreign investors and were built between 2003 and 2013. The research employs a difference-in-differences modeling approach to compare local regions around the corridor areas, as well as those further away that are not directly targeted by the corridor investments. The projects are analyzed for more than 10 years after the start of construction. Projects that are built in regions with an existing economic baseline are associated with the most growth. New built projects that are in landlocked countries are associated with less growth.

This study presents two novel strategies for identifying outlier parcels in urban deliveries, focusing on cost and environmental impact. With the growth of Business-to-Customer (B2C) e-commerce, the logistics industry faces increased pressure to improve last-mile delivery (LMD) efficiency. Outlier parcels, characterized by higher costs and emissions, pose significant challenges to logistics service providers (LSPs). Using the Marginal Cost Method and the COFRET method, this study develops cost-based and emission-based identification strategies to help LSPs address inefficiencies. The elbow point method is introduced to set objective thresholds for identifying outliers. Results from simulations in the South Holland region reveal significant differences in outlier parcel distributions across various carriers. Additionally, a sensitivity analysis assesses the impact of carbon pricing on parcel identification. The findings provide actionable insights for LSPs to optimize delivery operations and achieve sustainability goals. Future work will explore the application of these methods in different regions and consider alternative delivery methods for handling outlier parcels.

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.

In response to the inefficiencies in road freight transport, this research proposes an innovative collaboration index for carriers. The framework involves a collaboration decision tree that determines four collaboration modes at the trip level, which are then aggregated into a carrier-level evaluation using heuristic local optimization. This results in a collaboration index that quantitatively assesses the potential for collaboration among carriers based on operational data, filling a gap in partner selection processes. The study reveals that participants in horizontal collaboration achieved an average cost-saving rate of 14.4%, with the top 25% reaching 22.12%. While geographical similarity moderately correlates with the index, vehicle type does not show a linear relationship.
The significance of this work lies in its unique collaboration index formulated from operational data, which not only facilitates the identification of collaboration potentials but also provides insights for collaborative partner selection strategies.

In this research, a step by step approach is followed to find the root cause of the lagging adoption of an automated planning system by equipment planners. This is done by exploring the Technology Acceptance Model by Davis, from which a conceptual framework is established. This framework is central to a survey conducted under equipment planners globally. The results of this survey, and follow-up interviews, allow to identify the root cause of the lagging adoption of the new system. Furthermore, these results are used to draft a road map to improve the adoption rate.

This thesis investigates a container drayage problem involving terminals, depots, and shippers. Each terminal operates a homogeneous fleet of trucks that start and end at their respective terminals, carrying either one 40ft or two 20ft containers, and can make multiple trips within a single planning horizon. Terminals handle full containers and maintain a limited stock of empty containers, while depots provide additional empty containers. Shippers have specific time windows for service, and requests include sending and receiving empty or full containers to designated terminals, with street turns of empty containers possible. An Adaptive Large Neighborhood Search (ALNS) algorithm is implemented to address this problem, outperforming traditional methods like CPLEX. Extensive computational experiments validate the ALNS algorithm's efficacy, highlighting the logistical benefits of optimal depot placement and the impact of street turns, emphasizing the practical implications for container logistics.

In their Distribution Centre (DC) in Maasdam, FrieslandCampina (FC) uses a four-crew shift schedule to prepare all necessary orders for their clients, 24 hours of each Monday to Saturday. Their large automated warehouse is home to 10 000 pallet places, containing fresh dairy products. From here, orders are either prepared as full pallets, machine-picked layers or hand-picked "colli". In the last department especially, personnel cost is high relative to the throughput. Definitive picking deadlines are often ambiguous, posing challenges in job and personnel scheduling. The study goal is twofold. Firstly, to find out whether full knowledge of picking deadlines can contribute to a more efficient job, and so, shift schedule. Secondly, to offer insight for a trade-off between shift types to absorb workload. To reach this study goal, a Shift Minimisation Personnel Task Scheduling Problem (Krishnamoorthy et. al., 2012) and a Bin Packing Problem (Paquay et. al., 2014) were combined and tailored to fit the scheduling problem at FC's DC. In three weekly scenarios, the MILP model scheduled picking jobs in the least expensive shifts through a cost minimisation function. Two model configurations were used, one to prefer the shift between 09:00 and 17:00 (flex), and one to prefer either one of the 06:00-14:00 (morning) or the 14:00-22:00 (afternoon) shifts. Both model configurations inherently avoided the most expensive 22:00-06:00 (night) shift. Main findings include the possibility to absorb workload using the morning and afternoon shift and to avoid the night shift. Additionally, it was confirmed that insight in picking deadlines can contribute to an efficient personnel schedule a great deal.

The installation rate of offshore wind energy has to be quadrupled by 2030 to meet the green climate ambitions of European countries, but this growth is hindered by logistical challenges. Therefore, this study explores how two transportation and installation strategies, shuttling and feedering, affect the installation rate with the inclusion of manufacturing ports. Shuttling is where the installation vessel collects components itself at a port, and feedering is where the installation vessel remains offshore and gets supplied directly via feeder vessels. This is a novel approach as feedering and manufacturing ports are often not considered and production rates at manufacturing ports have not been considered at all. A rolling horizon simulation model is developed, which uses a Markov simulation model for weather forecasting, with a 72.92% forecast accuracy over two weeks, and a greedy algorithm for transportation and installation optimization. Results indicate that accurate initial buffer calculations, depending on the production rate at the manufacturing ports and project-dependent characteristics, can increase the installation rate significantly for either strategy. Shuttling becomes more efficient than feedering if the distance between the manufacturing ports and the offshore wind farm is too large or the size of the feeder vessels is too small. Feedering is more efficient in all other circumstances and, on average, results in a 9.2% higher installation rate and reduces project duration by 29 days compared to shuttling.

This research aims to enhance the understanding and quantification of logistics in the Netherlands by identifying factors influencing freight activities in urban areas. Using a linear regression model developed from BasGoed freight tour data, CBS location data, and Dutch zonal plan data, the study integrates 26 variables to predict freight trip generation and freight generation. The main explaining factors include the surface area of distribution centers, employment in various sectors, the presence of rail or inland waterway terminals, urban density, and the surface area of business spaces. The model, validated through statistical tests and literature comparisons, provides reliable predictions, with employment influencing freight production and surface area affecting freight attraction.

A case study in the Utrecht A12 zone demonstrated the practical relevance of the model, offering insights for policymakers to design logistics-friendly urban environments. Limitations include simplified statistical procedures and reduced accuracy in seaport and transhipment areas. Despite these constraints, the model delivers valuable predictions for various spatial scenarios.

Future research should explore model performance using CBS XML-microdata, consider non-linear relationships, and examine variables for shipment size and vehicle type. Addressing causality and endogeneity with instrumental variables is also recommended. This study contributes significantly to both scientific and societal domains, providing a robust model for freight trip generation and freight generation, and guiding municipalities in effective spatial planning and decision-making for sustainable urban development.