Public transport timetables are frequently redeveloped to fit within operational, structural and contractual constraints. In specific, passenger-oriented timetables are usually evaluated with a special focus on travel time, waiting time, transfers, reliability and fare cost. However, there are activities in people’s daily lives that have a fixed and recurring start time. This means that individual passengers care about the time they arrive at a specific location.

Therefore, a public transport service can perform well according conventional measures, while still not accommodating a passenger’s ideal arrival time. This is specifically important for locations to which many users converge, resulting in recurring schedules (e.g. schools, higher education institutions, shift-based work places, transport hubs). Despite these ideal arrival times, existing timetable design and evaluation methods do not consider passenger-oriented synchronization. Rather, they predominantly focus on time durations, reliability, or transfer synchronization. Hence, this thesis aims to develop a generic schedule alignment indicator framework. This framework evaluates how well scheduled vehicles align with passenger-desired arrival time window at relevant activity locations. To obtain this goal, the concepts of schedule delay (i.e. the divergence between desired and actual arrival times) are operationalized to be used in an applied public transport context.
The main research question that governs this process is: What constitutes a generic evaluation framework to evaluate timetable quality with respect to passenger-desired arrival times at activity locations?

The framework utilizes different stages to convert data inputs into a measurable timetable indicator. First, relevant activity locations are selected by connecting them to nearby stops using walking accessibility, and analyzing the demand of those stops. Then, their relevance is determined using the total amount of location users, combined with the mode share of public transport. When it comes to the construction of the desired arrival time windows, activity start times need to be obtained or estimated. In order to construct a window, the earliest, ideal, and latest arrival moment are defined relative to the activity start time. From these points, a triangular scoring function is constructed, with a score of 1 at the ideal moment and 0 at/outside the boundaries of the window. The arrival windows are transferred to the public transport stops using estimated egress times. To evaluate a timetable, the scheduled arrival times are scored using the scoring functions. With an impact-based weight, different score aggregations can be done according to the desired insights (e.g. on stop, route or network-wide level). In case of data unavailabilities, the framework relies on fallback methods to change individual steps in a modular way.
... Read more

Cities are increasingly pressured by rapid urbanisation, climate change, and congested transport networks, intensifying the need for sustainable mobility solutions that prioritise walking and cycling. Designing effective pedestrian-friendly interventions requires reliable, city-scale models of pedestrian activity. However, most existing transport models remain car-oriented or focus on microscopic pedestrian dynamics. Macroscopic pedestrian models are often highly data-intensive, rely on household surveys or extensive pedestrian counts, and are poorly suited to the Dutch context, where such data are limited. This creates a clear gap: the need for a practical, open-data-based, building-level pedestrian trip-generation model for Dutch cities.

To address this gap, this study develops a building-level pedestrian trip-generation model with fine spatial and temporal resolution, tailored to Dutch conditions and relying exclusively on widely available open data. The central research question is: To what extent can an existing pedestrian trip-generation model be adapted to reflect pedestrian trip-generation dynamics in the Dutch urban context?

The study adapts Sevtsuk’s Urban Network Analysis (UNA) framework into a Dutch-specific, building-level model referred to as BPT-Gen (Building-level Pedestrian Trip Generation). The UNA framework was selected because it offers a practical balance between spatial detail and data requirements while remaining transparent and reproducible. Several key adaptations are introduced to ensure applicability to Dutch cities.

First, buildings are identified and classified using the Dutch BAG dataset, supplemented with OpenStreetMap (OSM) data to capture land use, amenities, and public transport facilities. This enables a detailed and consistent representation of pedestrian trip origins and destinations at the building level.

Second, the derivation of building activity weights is modified. These weights represent the estimated number of unique daily users per building. Where detailed local data are available, weights are calculated directly; otherwise, Dutch building and occupancy standards or carefully selected proxy indicators are applied. This approach allows the model to remain operational across data-limited contexts while explicitly acknowledging uncertainty.

Building on the UNA framework, the model incorporates accessibility-adjusted activity weights. Each building is connected to the pedestrian network, and destination-specific Reach indices are calculated using a walking radius relevant to Dutch conditions. These indices adjust baseline activity weights to account for behavioural tendencies and destination attractiveness. Sensitivity analyses examine how assumptions regarding walking radius and normalisation methods influence model outcomes.

A major extension beyond the original UNA framework is the integration of temporal dynamics through hourly walking trip rates. Trip purposes derived from the Dutch ODiN travel survey are mapped to building types, and corresponding hourly origin–destination trip rates are applied. Combining these rates with accessibility-adjusted weights yields hourly, building-level pedestrian trip-generation estimates for the study area.

Model performance is assessed through face validation using observed pedestrian counts. Results show that the adapted model reproduces realistic spatial and temporal patterns for building types supported by reliable activity data, particularly housing and major train stations. Underprediction is observed for offices and leisure facilities, where activity weights rely on proxy indicators and where walking trips are likely underreported in ODiN. Sensitivity analyses confirm that trip-generation outcomes depend strongly on proxy selection and accessibility assumptions, revealing structural uncertainties.

Despite these limitations, the BPT-Gen model provides clear practical value. It identifies key generators of walking trips, peak periods, and accessibility-driven hotspots using only open data. By linking trip purposes to building types at hourly resolution, the framework fills an important gap in Dutch pedestrian modelling and offers a transparent foundation for future extensions. The findings demonstrate that open-data pedestrian models can provide meaningful planning insights while highlighting the need for improved activity indicators and further validation using street-level pedestrian flow data.
... Read more

Bicycle travel demand analysis remains relatively underexplored, yet it is becoming increasingly important for urban and campus planning. In the Netherlands—particularly within the TU Delft community—cycling is deeply embedded in daily life, creating a strong demand for supportive infrastructure. While congestion may not pose the same threat to bicycle transport as it does to motorized traffic, maintaining the performance and safety of the bicycle network is essential. A busy cycling network may not always be visible at a glance, yet it can present safety risks.

The primary aim of this research is to identify a modeling process and specifications that are compatible with the available data, while laying the groundwork for future improvements to the bicycle network, especially TU Delft Campus. This will help ensure the system remains adaptable and relevant for long-term planning.

To identify an appropriate modeling approach, an exploratory analysis of the data was conducted. A clear pattern emerged in bicycle traffic, characterized by short-interval fluctuations corresponding closely with lecture schedules. An additional notable observation is the occurrence of an average peak in bicycle traffic during midday. These findings support a dynamic analysis approach with a 5-minute interval.

Moreover, the model incorporates specialized variables defined by the study’s scope, focusing on trip generation and trip distribution within the established four-step modeling framework, specifically tailored for Origin-Destination (OD) matrix estimation in transportation engineering.

For trip generation, linear regression coupled with backward stepwise elimination via the Ordinary Least Squares (OLS) method was employed to identify significant predictors. For trip distribution, the Iterative Proportional Fitting (IPF) method was utilized. This approach was chosen based on the assumption that impedance is minimal for short-distance travel, a scenario particularly relevant within the TU Delft campus context.

Ultimately, this methodology provides a flexible and responsive framework tailored to the specific transportation dynamics at TU Delft, producing valuable insights for optimizing bicycle network planning.

The developed model is relatively simple but exhibits several shortcomings. One significant limitation is related to data collection, as the available data lack the temporal resolution necessary to fully capture the dynamic travel patterns targeted by the model. Additionally, the linear regression approach used for modeling trip production and attraction yielded unsatisfactory results, with $R^2$ values below 0.5. Another issue is potential underfitting, as indicated by the improved explanatory power of the model when trained on smaller datasets. Validation using RMSE and comparative plots of modeled versus actual flows further confirms that substantial improvement is needed in the model’s reliability and predictive capability.

The trip distribution process, conducted using the Iterative Proportional Fitting (IPF) method, reveals additional areas for improvement. The OD matrix underestimated total production by two bicycles in a 5-minute interval. Although seemingly small, this discrepancy underscores the necessity for more robust input data and methodological refinements. Additionally, direct validation of the OD matrix is crucial to enhance accuracy and reliability in representing actual travel flows.

Despite the shortcomings, the framework provides a balance between interpretability and flexibility, enabling both accurate representation of observed travel behavior and ease of scenario testing—making it a practical tool for supporting data-driven mobility planning and policy evaluation on campus.
... Read more

This study investigates the impact of lowered lock capacity on lock demand shifts within the Weurt- Grave system, a critical link in the Southeast inland waterway corridor of the Netherlands. It answers the following research question: What impact does a reduction in capacity, caused by low water levels, have on alternative locks within the inland waterway system, with a focus on the Weurt- Grave system? Understanding the interactions is important for improving waterway infrastructure planning. Lock ca-pacity constraints caused by extreme water levels are becoming more likely to occur due to climate change. This increases the importance of being able to predict the effects these disruptions will have on the larger waterway system. This research shows from historical data how lock capacity decreases can affect other locks. This helps give insight in the scope that might need to be assessed when researching a single lock that is part of a larger system. During dry periods, low water levels in the Waal limit lock Weurt’s capacity, redirecting larger vessels to lock Grave. This redirection alters waiting times, fleet composition, and traffic patterns within the system. It shows the system’s sensitivity to environmental conditions. Using lock usage and water level data from Rijkswaterstaat, the study applies statistical methods, including logit regression, time series analysis, and moving averages, to quantify water level impacts on lock utilization and traffic patterns. The findings reveal redistributions in vessel traffic and load, primarily driven by vessel draught, origin- destination patterns. While the system shows adaptability, capacity constraints at Grave raise concerns about infrastructure durability under increased demand. Furthermore, the research identifies limitations in existing traffic models, such as SIVAK and BIVAS, which fail to capture dynamic interdependencies between locks under changing environmental effects. This study shows that lock capacity constraints should not be viewed as an isolated problem. These constraints could be causing network wide effects, changing traffic flow and fleet compositions at other points in the system. By revealing how vessel behaviour responds to changing capacities, the research emphasizes the necessity of network wide approaches to lock modelling. Knowing of these possible system-wide effects is essential for modelling waterway infrastructure that need to handle the results of climate change and changing transport demands. This will improve the resilience and efficiency of the inland waterway network. ... Read more

This study explores the potential role of e-scooters in supporting “first-mile” travel to train stations in the Netherlands, a critical stage in daily commuting for many Dutch travellers. In a nation where bicycles are a primary mode of transport—especially for trips under 5 kilometres—the introduction of e-scooters offers an alternative with similar benefits: environmental friendliness, cost-effectiveness, and ease of manoeuvrability. With their compact size and accessibility through sharing services, e-scooters can complement the existing bicycle culture, promising an appealing option for urban mobility.

The research focused on examining the role of e-scooters in “first-mile” travel through a stated preference choice experiment, given that e-scooters were newly legalised. A sample of 156 participants responded to six hypothetical travel scenarios, where they chose between familiar modes like walking, cycling, e-bikes, and e-scooters based on different factors such as carry-on ticket costs, parking times, and travel times. These scenarios were designed to reflect typical Dutch commuting choices while testing the appeal of e-scooters as a new transport option.

Analysis of the data using a Multinomial Logit (MNL) model yielded several insights into travel behaviour. There was a clear baseline preference for familiar modes, with participants showing a higher likelihood of choosing options like cycling or walking. Travel cost and time emerged as significant decision-making factors, with higher costs and longer times deterring mode choice. Notably, gender differences surfaced in this context, with males displaying greater tolerance for longer bike trips than females, indicating variations in time-cost sensitivities across demographics.

Experience with transport modes also influenced choices; prior e-scooter use was positively correlated with selecting this option again, suggesting that familiarity can increase comfort and confidence in choosing new modes. Established travel habits showed a strong impact on choice consistency, with individuals who typically walked or cycled to stations likely to maintain these habits, reinforcing the role of routine in travel preferences.

To illustrate the findings, a simulation modelled two scenarios. In the “Extreme Low” scenario, where e-scooter costs were low and parking time was minimal, walking was favoured for short distances (under 0.5 km), while cycling became dominant as distance increased. In contrast, the “Extreme High” scenario, with no fees and longer parking times, significantly boosted e-scooter attractiveness, particularly beyond 1.5 km, where it surpassed both walking and cycling. This outcome underscored the sensitivity of user preferences to economic factors, with e-scooters emerging as a highly competitive option when cost barriers were removed.

Overall, this study highlights the substantial potential for e-scooters to influence first-mile travel choices, particularly when costs are favourable and convenience is enhanced. By offering a viable alternative for medium-distance trips, e-scooters could reshape first-mile mobility in the Netherlands, complementing the well-established cycling culture and contributing to sustainable urban transport. ... Read more

The increasing prevalence of micromobility vehicles in urban environments has raised concerns about safety in shared cycling spaces. This study examines the overtaking behavior of e-scooter and e-bike riders to inform traffic management strategies and infrastructure development. A controlled experiment was conducted using strategically placed cameras to track vehicle trajectories and inertial measurement units (IMUs) to capture roll data. Extensive data processing ensured accuracy and synchronization of trajectory and IMU information. Key findings reveal that e-bikes overtaking e-scooters initiate maneuvers from greater distances but maintain smaller lateral distances compared to e-bikes overtaking e-bikes.
Lateral position differences showed a stronger correlation with speed difference than longitudinal position differences. The highest roll rates and angles occurred during the overtaking phase. Pre-overtaking, higher roll rates and angles were observed when e-bikes overtook other e-bikes, indicating greater control adjustments. No significant gender differences were found in overtaking behavior. However, in non-interactive scenarios, male e-scooter riders traveled at higher speeds than females, while no gender differences were observed among e-bike riders. These results provide insights into the complex interactions between different types of micromobility vehicles during overtaking maneuvers. The findings underscore the need for targeted safety interventions and infrastructure improvements to mitigate risks associated with shared cycling spaces, ensuring safer coexistence of micromobility users and conventional cyclists in urban environments. ... Read more

This research aims to understand the influencing attributes of passengers and station layout elements on transfer walking times for metro transfers. Little research has been performed to include passenger-related attributes to model transfer walking time besides station layout elements. Through a literature review, the effect of gender, luggage size, group size and level of crowding are potential attributes. Furthermore, the vertical transport mode choice, including the lift, the waiting condition to board a vertical transport mode and the alighting location, are also part of the influential attributes besides the transfer length. The walking time and passenger characteristics have been collected through a covert observation. In the data analysis, the effect of group size, vertical transport mode choice, waiting condition to board and the alighting location significantly impact the transfer walking time and the walking time on a transfer segment level. These attributes have been captured in walking time and passing speed models for various transfer segment types. The walking time models can predict a lower, mean and upper bound of the walking time for each combination of attributes. The case study for the walking time collection was metro station Beurs, Rotterdam. ... Read more

This research provides insight into the most effective approach for implementing shared micromobility services in towns and suburban communities, considering the preferences and requirements of key stakeholders, such as users, non-users, shared micromobility providers, public transport operators and municipalities or other government agencies. The scope of the study is limited to sharing of bicycles, e-bikes and mopeds. The results will allow all municipalities within the same province, or even within the entire country, to ensure a unified application with a solid business plan and a greater chance of success.

An example of a promising service involves the application of shared bicycles with hubs at all key locations in the service area and various hubs in the neighbourhood (back-to-many). Anticipating on the requests of residents and businesses. The service is connected to all municipalities in the area, with an emphasis on connecting at least one large town or city. If this collaboration is established, shared e-bikes are introduced, primarily focused on distances exceeding 3 kilometres. Finally, the system can be complemented with shared mopeds, limiting their use to a few hubs per municipality. This allows for effective control of usage and minimizes disturbances. Limiting the number of locations makes it possible to create physical parking facilities for mopeds, which contributes to mitigating inconvenience. However, every situation is unique, this thesis provides guidance on the approach to be considered for different scenarios. ... Read more

With an increase in interest in putting people on bikes instead of driving vehicles for day-to-day transport, new regulations are being implemented to make roadways safer for cyclists while also encouraging cycling behavior among various segments of the population. The perception of safety has been proven to play a significant role in the decision to adopt active forms of transportation. The significance of this feeling is underlined even more when it comes to vulnerable road users, in this case, cycling children. This study seeks to provide a better knowledge of the road conditions that increase the perception of road safety in order to establish a safe and encouraging cycling environment for adolescents cycling to school. The preferences of different categories of the population toward various measures are explored using the stated preferences experiment. Furthermore, with the adoption of the new lowering of the speed limit policy, the attitude toward a growth in the usage of bicycles is done. According to the findings of this study, different sectors of the public have distinct perceptions and preferences toward the numerous aspects shown on the highways. The presence of separated bike lanes, a speed restriction of 30 km/h, and a priority for crossing rule for cyclists are discovered to be components that all groups in the sample strongly prefer. Parking on the country road is limited to the groupings of parents and road designers. Finally, the presence of speed displays is discovered to be negligible for all subgroups in the sample population as a measure that has no influence on making roads safer for children to cycle on. Furthermore, it has been discovered that parents who bring their children to schools that
are now located near unsafe GOW50 roads have a more positive attitude toward cycling following the implementation of the speed limit policy. More aspects that were discussed in the GOW30 debate could be included in a future study to gain a better understanding of the impact of the various elements on providing a safe cycling environment for children. Furthermore, a recommendation is made to examine the behavior of parents in other locations that are also in close proximity to hazardous GOW50 roads and to incorporate the features of other respondents in order to corroborate the findings of this study. ... Read more

Recent years have seen a rise in computation power, storage and opensource data has become more accessible. This results in new opportunities for microsimulation models. Most of these models require a realistic population but the data for this is usually not available. Therefore, population synthesis is used to generate the statistical equivalent of the real population. Since the majority of trips, tours and activity schedules are centered around the home end. It is also important to not only generate a synthetic population but to also attach these households to actual houses. A geodata source is needed for this and OpenStreetMap can be a viable data source. This research focuses on developing a proof of concept of a method for population synthesis at the fine scale of neighborhoods and using OpenStreetMap data to add spatial units to this population. The methodology is developed through literature and implementation in a case study. ... Read more

Promoting cycling is an effective way to realize a more sustainable transportation system and healthier living environment. Authorities have the means to play an important role in this with the formulation of policies and the implementation of cycling interventions. However, they lack the knowledge on how to determine the effectiveness of cycling policy. The findings of this research support practice with insights from the literature on policy monitoring for cycling on how to better monitor the effectiveness of cycling interventions. A theoretically-underpinned Monitoring Framework is developed that shows what the necessary elements are to monitor policy, what different aspects of policy can be monitored, and how policy monitoring is set in the process of policymaking. The framework is used to systematically compare research findings from a literature scan and case study. In the case study, policy documents have been analysed and interviews with policymakers were conducted to study current policy monitoring practices at two regional authorities in the Netherlands: the regional transport authority, Vervoerregio Amsterdam (VRA) and the Province of Noord-Holland (PNH). Conclusions are made about knowledge gaps in the monitoring of cycling policy in practice which relate to the three components of the Monitoring Framework. Future research recommendations are made on how the Monitoring Framework can be expanded by zooming in and out from the scope of this research. Recommendations for practice are made on how to better embed policy monitoring in policymaking. ... Read more

The COVID-19 pandemic has changed people's lives in greater or lesser degree since the outbreak back in December 2019 in Wuhan, China. Due to the high transmissibility of the virus and with the aim of preventing the virus from continue spreading, several measures have been applied during the last two years, which now have become part of people's daily life. One of these measures is physical distancing, which has been proven to be an effective way to reduce the transmission risk. The objective of this research is to understand how the walking behaviour of pedestrians has changed during COVID-19 because of physical distancing at bidirectional flows in locations where people might encounter difficulties for complying with this measure, and conduct a calibration of the pedestrian simulation model Pedestrian Dynamics (PD) to improve its accuracy for reproducing such behaviour. As a result, this research aims at determining the capability of the model to reproduce the walking behaviour during the pandemic. The results show that the walking behaviour has changed over the course of the pandemic, since in 2020 and 2021 pedestrians kept respectively larger and shorter distances than in 2019. Moreover, a larger variability in velocity and higher effort to traverse the measurement area suggest changes in the awareness of their surroundings during the pandemic. The calibration results show that the relaxation time is the most relevant parameter as the accuracy of the model significantly changes when values lower than its default are used, which in turn indicates a faster reaction of pedestrians. In conclusion, a calibrated model considering one parameter would not yield accurate predictions if a general usage for all COVID-19 scenarios is intended. Finally, different types of behaviour are recommended to be analysed to obtain insight into the capability of the model to reproduce them and moreover, the behavioural assessment at strategic and tactical level are adviced, since this research focused on the impact of physical distancing on the behaviour at operational level. ... Read more

High vessel traffic densities, restricted water areas, and low manoeuvrability are common factors that form constraints for the navigating vessels in port areas. While navigating through these areas, vessels get assisted by the Vessel Traffic Service (VTS). VTS operators (VTSO) are monitoring the vessels in ports and provide the navigating skippers with the required information to ensure the safety and efficiency of the vessel traffic in the port. Information is communicated using very high frequency (VHF) radios on the corresponding VHF-channels of the sectors. In busy sectors in the port, these VHF-channels are very crowded, which results in unclear situations and high workloads for the VTSOs. The Rotterdam Port Authority is currently investigating measures to reduce VHF-traffic. To define effective measures, it is required to understand by which factors the VHF-communications are triggered. In this thesis, a quantitative method has been developed to investigate whether VHF-communications are caused by certain properties of a fairway system. The method is based on analysing VHF-traffic and identify factors that lead to an increase in these communications. The main component is digitizing the VHF-communications, which is achieved by converting spoken VHF-audio bands to text, then dividing this text into different conversations and finally classifying each phrase according to place, time and content. Subsequently, the spatial and temporal characteristics of the VHF-communications are linked to spatial and temporal characteristics of a property of the fairway system, to analyse whether the occurrence of a particular form of VHF-communication is related to the occurrence of that particular property of the waterway system. The method can be applied to investigate the individual influence of several fairway properties, but in this research the focus is on analysing to what extent the fairway property “blocked view” leads to the development of VHF-communications. To investigate whether VHF-communications are related to the occurrence of visual obstructions, a method is developed to estimate the visibility for varying locations and vessel parameters. ... Read more

Travel time is an integral part of all human activities. In transport planning, travel time is one of the most widely used and significant parameters in the design of transport systems. However, people's ability to estimate travel time and time, in general, is limited. The aim of this thesis is to broaden the knowledge on the perception of travel time of active modes. More specifically, this thesis's objective is to provide insight into the deviation of perceived travel time to the actual trip duration of cycling and walking trips and to identify the factors that influence it. Individual characteristics and emotional state, trip characteristics, and external conditions such as the weather are considered. To reach the research aim, data are collected on the subjective and objective travel time and on the possible determinants. A mobile application is developed to serve as the single data collection tool. The app incorporates a location (GPS) data source and multiple surveys. The trip duration, which is reported by the survey respondents right after the completion of a trip, is considered as the subjective travel time. Data are collected for two weeks in the Netherlands in June 2020. An analysis is performed on the obtained dataset in terms of descriptive statistics and regression analysis. The former reveals that, on average, people underestimate the duration of trips conducted by bike or on feet. Via linear and logistic regression models, the decisive factors of travel time misperception are pointed out. Disappointing weather, high physical effort demand, and compulsory trip purposes prolong the perceived trip duration. On the contrary, the usage of an information source for planning a trip leads to a shorter subjective travel time. The overall underestimation of travel time by active modes indicates that the perceived disutility due to travelling could also be lower than the objectively measured one when travelling by these modes. The findings of the study are discussed in regards to recommendations to practice and future research. Furthermore, the thesis provides suggestions for further usages of mobile applications for the collection of data on travel time distortion and on travel experience, in general. ... Read more

To intrigue more people to cycle, it is essential to assess the current design of the cycling infrastructure to guarantee the efficiency and safety for the users. A credible simulation of cyclists is helpful to evaluate different designs of infrastructure and point out the problem. A valid simulation model requires proper verification and validation. However, there was hitherto no standard procedure to conduct model verification, and limited studies were found related to the verification of cyclist simulation models. To fill this gap, this study develops a verification framework for traffic simulation models based on the methodologies used in the literature. It also briefly discusses the difference to verify a cyclist simulation model. The framework is assessed by applying to a case study of the cyclist-queue-formation simulation model. The results prove the applicability of the framework to verify a cyclist simulation model. The organised structure of the framework provides a systematic measure for the verification of traffic simulation models. Furthermore, the cyclist-queue-formation simulation model is verified within the study area, where its conceptual model is proposed. The room for improvement of the simulation model regarding the relatively slow correction behaviour in steering and pedalling of the cyclists was also indicated. ... Read more

With the upsurge of disruptive trends such as digitalisation and automation, organisations have become more vulnerable to the consequences of these disruptive trends. The Harbourmaster is responsible for facilitating a safe and efficient port-call-process. Digitalisation and automation might lead to opportunities and threats in the port-call-process the Harbourmaster cannot for see. As a result, the need for more resilient strategies to respond to these trends has become paramount. This research aims on filling the gaps in literature, by developing a strategy roadmap for the Harbourmaster to manage uncertainties in the port-call-process due to digitalisation and automation. By doing so, it provides a structured approach to examine the impact of digitalisation and automation in the port-call-process.

Digitalisation and automation affect the capabilities of the Harbourmaster's organisation with respect to long-term-objectives, the impact on the current port-call-process and the applications and the impact on staff qualifications. Adjustments to the Dynamic Adaptive Policy Pathways framework have been made to fit it into the objectives of this qualitative research. Resulting in thirteen effective actions that contribute to long-term-objectives to respond to the uncertainties (both opportunities and threats) of digitalisation and automation in different futures of the port-call-process. All actions focus one of the areas affected by digitalisation and automation. The resilient strategy roadmap visualises correlations between the actions.

Concluded is that a qualitative application of the DAPP can be used to design resilient strategies for the Harbourmaster. The resilient strategy roadmap provides the Harbourmaster with a guideline that is useful in uncertain environments due to digitalisation and automation. It can be discussed if other cases would experience the similar results when applying the DAPP framework in a qualitative manner. Furthermore, different set-up of methodologies is expected to result in a more valid set of actions. In addition, the impact of autonomous vessels is recommended to be further investigated since this is expected to have a radical impact on the port-call-process. ... Read more

Knowing 'where I am' is always essential and a prior to answer for a moving vehicle. Among numerous onboard sensors, a GNSS receiver for single-frequency Precise Point Positioning and camera are competitive due to the fairly lower cost and the potential to provide a lot of useful information.

However, due to the degraded GNSS solution performance in city valleys, a tight integration is considered combining the two sensors at the observation level, ie. processing the GNSS ranges and the vision measurements in the image of the camera. The availability of High Definition Maps (HD Maps) aids vehicle positioning by providing extra information on the environment. In this project, landmark positions are retrieved through vision and the HD map, and can complement GNSS in city valleys. Additionally, the project focuses on building the mathematical model for the integration of observed landmark position (using a single camera, considering the ease of implementation and cost) and GNSS measurements, analyzing the performance as well as the feasibility for vehicle positioning. The project emphasizes the feasibility study of the proposed mathematical model, which is flexible and capable of using all available input automatedly, and providing a position solution with the best precision.

The uncertainty in the available landmark positions (for instance errors in the HD maps) is handled in two different ways: one is to include the landmark position coordinates as measurements into the model, the other one projects the uncertainty onto the measurements in the camera image. The latter method turns out to be much more efficient. To integrate vision and GNSS measurements, a conversion between an ECEF (earth-centered, earth-fixed coordinate frame), typically used for GNSS, and a world coordinate frame for the camera measurements, is required. A position offset between the GNSS antenna and camera is considered, since the camera lens center does not coincide with the GNSS antenna center. In the simulation and experiment, an extended integration is also presented and discussed which leaves out the position offset, for instance when the GNSS antenna is very close to the camera, which can further improve the redundancy and lower the computational load.

From the simulation and experiment, we conclude that the integration model is able to produce a position solution when one of the sensors is unable to produce a position solution and the other one still can; the extended integration model is able to produce a position solution even when both sensors individually fail to produce a position solution. Among these scenarios, the one when GNSS fails and vision operates, the integration model can produce a position solution within a quarter of a meter in local horizontal coordinates, and the GNSS measurements do not contribute much to the position solution. Compared to the integration model, the extended integration improves the model by reducing or eliminating the (typically heavy) correlation between the estimates, in particular those for the camera-antenna position offset, the GNSS receiver clock error and the vertical coordinate. Under the same scenarios, the extended integration improves the standard deviation in vertical coordinate and receiver clock error, within a quarter of a meter and one-third of a meter respectively.

Further study is recommended in the direction of applying full image processing procedures to obtain more realistic vision measurements, to include GNSS carrier phase observations to replace the current GNSS positioning based on Precise Point Positioning, in order to have a position solution of similar quality as the vision part. The dimension of system gets larger when carrier phase measurements (phase ambiguities) are added as well as two additional rotations for a camera; the extra rotations introduce a significant amount of nonlinearity in the model. A larger model with increased nonlinearity may call for an alternative model formulation.
... Read more

Decision- and policymakers responsible for the coastal zone aim at combining measures against for instance long-term erosion, with measures that have a positive social, economic impact on the region. Recreational beach usage has a large social economic impact on a region and therefore quantification of the recreational beach usage can provide information on the social, economic situation in a region. In the Netherlands beach usage quantification is mostly performed by manual counting during a limited number of days in the field and this limits the spatial and temporal resolution. The objective of this study is to develop and test a method for accurate, robust and automatic monitoring of the spatial and temporal distribution of the number of beach users on the Dutch coast.
Multiple approaches to quantify the number of persons in an area are reviewed. Comparison of the reviewed approaches showed that a first distinction can be made between methods using a visual light (camera-) sensor and other methods based on the use of Bluetooth, Wifi, GPS-location (all related to phones) and LiDAR. Based on a literature review the visual light sensor is decided upon to best suit beach user quantification. Within the visual light approach a second distinction is made between methods based on the difference in pixel intensity, a method based on variance of pixel intensities over multiple frames and a method using oversegmentation combined with a machine learning framework for classification.
The difference in pixel intensity method is observed most often in literature, but has limitations in conditions that are concerned typical for the Dutch coast (e.g. clouds). The method based on the variance of pixel intensities and the method using oversegmentation are possibilities to overcome the problems described in the studies on the pixel intensity method. Based on preliminary tests the oversegmented machine learning approach is selected because it does not require beach users to move to be detected. Moreover, single snapshots can be evaluated which require a limited data infrastructure in-situ and this is considered advantageous regarding the ease of implementation and cost effectiveness.
The oversegmented machine learning approach divides images into small regions of similar pixels based on pixel gradients. The regions are called superpixels and superpixels can be characterised by significantly more features than the conventional r,g,b relations corresponding to regular pixels evaluated in the differences in pixel intensity method. The availability of an increased number of features provides more options to distinguish between classes during classification and this can be advantageous in difficult (e.g. cloudy) conditions. Classified beach user superpixels can represent multiple beach users due to for instance occlusion and therefore a regression relation between classified beach user superpixels and a manually counted ground truth is determined for conversion of classified beach user superpixels to the number of beach users. Hence, the oversegmented machine learning method for quantification of beach occupation combines an oversegmented classification model to classify superpixels into classes (e.g. beach user and sand) and a regression model to convert classified beach user superpixels to beach users. 
The oversegmented machine learning method has previously been evaluated in the study of Hoonhout et al. (2015) for the classification of coastal images into the classes 'water', 'sand', 'objects', 'vegetation' and 'sky' and this led to the open-source toolbox Flamingo (Hoonhout and Radermacher, 2014b). The current study adapts and develops the Flamingo toolbox for the quantification of beach occupation. The impact of changing parameters of the existing toolbox on the oversegmented classification model are evaluated to obtain insight in the parameters that have to be changed to apply the toolbox to the quantification of beach occupation. The influence of the parameters: class aggregation, measures to take account for imbalances in the dataset, regularisation, number of images in the training dataset, image enhancement, addition of articial channels to enable more (new) features and the required number of features are reviewed. Especially changes in the parameters class aggregation, number of images in the training dataset and articial channels affect the overall model performance. The effect on the overall model performance of measures to account for imbalances in the dataset is limited. However, these measures can change the relative distribution of precision and recall corresponding to the false negative and positive rates respectively. The final classification model is trained and validated with a dataset containing 76 manually annotated images, default undersampling to account for the imbalance in the dataset and added articial channels. A 4-class model with classes beach users, sand water and objects proved to be the best performing class aggregation.
The classified beach user superpixels are converted into a number of beach users with a regression model obtained by fitting a second order polynomial regression line to the classified beach user superpixels of the training images and the corresponding manually counted ground truth. Evaluation of the fit shows that the oversegmented machine learning method is a suitable method for quantification of beach occupation indicated by a R2 of 0.92. The regression model is validated by application of the combined  oversegmented classification- and regression models on a new and 'unseen' dataset of 80 images. Validation shows that the regression model is applicable on images that are not used during development of the model (R2=0.87) and this moreover confirms the suitability of the oversegmented machine learning method. Analysis of the largest errors showed that especially unoccupied beach stretchers and images captured by an unclean lens limit the performance of the oversegmented machine learning method.
The developed oversegmented machine learning method is benchmarked against one of the differences in pixel intensity methods representing the current state-of-the-art. The benchmark shows that the oversegmented machine learning method (R2=0.87) has a higher performance on the evaluated evalidation dataset compared to the method representing the current state-of-the-art (R2=0.76). The difference in performance indicates that the newly developed method is more suitable to the varying conditions associated with the Dutch coast.
Tests of the oversegmented machine learning model on a different camera station than was used for training of the oversegmented classification model, did not lead to satisfactory results. This indicates that the current approach for application on camera stations not used during training is not suitable. Therefore, at this point, the oversegmented machine learning method lacks robustness with respect to the performance on multiple different camera stations. A number of possible causes for the limitedperformance are treated and provide recommendations for further research. 
The presented oversegmented machine learning method, despite its limitations, provides an opportunity to quantify beach occupation with a high temporal and spatial resolution in variable (weather) conditions that are known to limit the performance of the current state-of-the-art methods and are typical for the Dutch coast. The method, therefore, enables the possibility to monitor locations in conditions that with the current state-of-the-art would be diffcult to monitor. ... Read more