R. van Egmond
info
Please Note
<p>This page displays the records of the person named above and is not linked to a unique person identifier. This record may need to be merged to a profile.</p>
14 records found
1
Doctoral thesis
(2025)
-
Nicolas F. Chaves-de-Plaza, K.A. Hildebrandt, R. van Egmond, A. Vilanova Bartroli
Radiotherapy (RT) is a widespread and effective technique to treat cancers by killing cancerous cells with rays of radiation. Building upon advances in image guidance and dose delivery technology like Proton Therapy, Adaptive RT promises more effective tumor decimation and a reduction of the incidence and severity of side effects. Unfortunately, the clinical implementation of adaptive workflows is challenging due to their resource-intensive nature. Therefore, their successful adoption lingers on overcoming several bottlenecks in the treatment planning process.
In this dissertation, we focus on methods used for the image segmentation or contouring step, which allows the localization of the anatomical structures required for dose optimization and evaluation. Until recently, clinicians had to manually delineate dozens of organs-at-risk and target volumes across hundreds of slices of the patient’s three-dimensional images. A process that is extremely time-consuming. The advent of deep learning-based artificial intelligence (AI) has changed the landscape: a modern auto-segmentation AI can produce segmentations for most of a patient’s anatomy in minutes.
Despite increasing automation in the segmentation process, it remains time and resource-intensive. Due to the segmentations’ criticality for the patient’s outcome and the errors the AI will commit, clinicians must perform a quality assessment of the AI’s outputs. Depending on the case’s complexity, the duration of the quality assessment process can negate the time gains auto-segmentation tools bring.
Deep ensemble AIs represent an advancement in medical image segmentation. Instead of providing a deterministic output, deep ensemble AIs produce a set of plausible candidates that aim to model inter-clinician annotation variability. Consensus segmentations obtained from ensembles tend to be more accurate and robust than the single-prediction deterministic counterpart. Nevertheless, by only using the consensus, a lot of potentially useful information is being discarded.
In this dissertation, we contribute to different phases of the segmentation quality assessment process. We characterize this process and introduce methods that leverage the raw outputs of deep ensemble AIs to support and speed up quality assessment tasks. The methods presented show new ways of analyzing and using ensembles in RT. Nevertheless, since these are relevant outside RT, we keep the presentation of the methods general and evaluate them in other application scenarios, such as the analysis of simulation ensembles or meteorological data.
Before fixing segmentation failures, clinicians must find them. This process can be time-consuming and fatiguing when failures are sparse and spread through the patient’s three-dimensional images. We present and evaluate a delineation error detection system, which guides clinicians to slices of three-dimensional images that contain potentially clinically relevant segmentation failures. We co-designed the DEDS with clinicians and refined it based on an observational study, which allowed us to characterize clinicians’ navigation patterns and the use of information sources like AI uncertainty and patients’ dose distributions. We evaluated the DEDS’ potential to speed up the QA process through a simulation study with a retrospective cohort of patients. Results indicate that speed-ups are the most significant when equipping the DEDS with information sources indicative of clinical priority, which prevents unnecessary edits.
Visual inspection of the segmentation ensemble permits understanding the main trends and detecting anomalies that might indicate segmentation failures. Using a spaghetti plot to visualize all ensemble members is straightforward but prone to clutter. Contour boxplots prevent clutter and extra complexity by distilling essential ensemble information, which permits more efficient ensemble inspection. Nevertheless, they are time-consuming to compute, reducing their practical value. We present Inclusion Depth for contour ensembles. Inclusion Depth yields per ensemble member centrality scores that allow characterizing the distribution of segmentation ensembles in terms of properties like the median, trimmed mean, confidence bands, and outliers. Compared to previous contour depth notions, Inclusion Depth is significantly faster, making it more applicable in practice for time-critical contexts like QA in adaptive RT. We show how Inclusion Depth permits creating contour boxplots for ensembles with hundreds of segmentations in seconds.
It is not uncommon for distinct representative shapes to co-occur within a contour ensemble. With ensembles created by clinicians, for instance, different institutions, training sessions, or experience levels can lead to distinct shapes (i.e., modes of variation) for the same structure. When trained on these data, deep ensemble AIs would yield similarly multimodal ensembles. In quality assessment, being able to extract these representatives would pave the way for new ensemble-based interactive segmentation workflows. Applying traditional contour depth notions to these multi-modal ensembles collapses the existing variation modes and can lead to uninformative centrality scores. To address this issue, we present the first framework for multi-modal contour depth, which also includes notable runtime improvements for depth computation. When used with Inclusion Depth, multi-modal contour depth permits clustering the different modes of variation and determining cluster-dependent scores that appropriately characterize the data. Variation modes can be then independently analyzed using uni-modal depth machinery like contour boxplots. xiii
The global perspective of contour depth methods, which consider the entire volume, may be insufficient when parts of the contours are noisy or when the resolution of the ensemble is too large to process within a reasonable time. Correlation clustering methods provide a solution by partitioning the spatial domain of the ensemble into highly correlated regions that can be used to localize analyses. Existing correlation clustering algorithms do not scale well as the resolution of the ensemble increases. We introduce the Local-to-Global Correlation Clustering (LoGCC) method, which partitions the ensemble’s spatial domain into coarser primitives, representing areas of consistent ensemble member behavior. Unlike previous correlation clustering methods, the proposed LoGCC achieves significantly faster runtimes by leveraging the ensemble’s spatial structure and decoupling computations into local and global steps. Like with Inclusion Depth, these speed gains enable LoGCC to analyze large datasets in time-critical fields such as adaptive radiotherapy (RT).
Throughout this dissertation, our approach focused on designing modular, flexible analysis methods applicable across different tasks and domains. We demonstrate how the delineation error detection system, multi-modal Inclusion Depth, and Local-to-Global Correlation Clustering support quality assessment in RT and extend to fields like meteorology. We also speculate on their potential as foundational elements for more complex workflows. For example, extracted modes of variation, which indicate representative shapes in the ensemble, could be repurposed as an interactive segmentation tool. Alternatively, consistent regions detected by correlation clustering could be used as building blocks to enable localized contour analysis and editing.
We hope the proposed contour ensemble visual analysis methods inspire the development of more efficient analysis workflows that harness ensembles’ power in RT and beyond. ...
In this dissertation, we focus on methods used for the image segmentation or contouring step, which allows the localization of the anatomical structures required for dose optimization and evaluation. Until recently, clinicians had to manually delineate dozens of organs-at-risk and target volumes across hundreds of slices of the patient’s three-dimensional images. A process that is extremely time-consuming. The advent of deep learning-based artificial intelligence (AI) has changed the landscape: a modern auto-segmentation AI can produce segmentations for most of a patient’s anatomy in minutes.
Despite increasing automation in the segmentation process, it remains time and resource-intensive. Due to the segmentations’ criticality for the patient’s outcome and the errors the AI will commit, clinicians must perform a quality assessment of the AI’s outputs. Depending on the case’s complexity, the duration of the quality assessment process can negate the time gains auto-segmentation tools bring.
Deep ensemble AIs represent an advancement in medical image segmentation. Instead of providing a deterministic output, deep ensemble AIs produce a set of plausible candidates that aim to model inter-clinician annotation variability. Consensus segmentations obtained from ensembles tend to be more accurate and robust than the single-prediction deterministic counterpart. Nevertheless, by only using the consensus, a lot of potentially useful information is being discarded.
In this dissertation, we contribute to different phases of the segmentation quality assessment process. We characterize this process and introduce methods that leverage the raw outputs of deep ensemble AIs to support and speed up quality assessment tasks. The methods presented show new ways of analyzing and using ensembles in RT. Nevertheless, since these are relevant outside RT, we keep the presentation of the methods general and evaluate them in other application scenarios, such as the analysis of simulation ensembles or meteorological data.
Before fixing segmentation failures, clinicians must find them. This process can be time-consuming and fatiguing when failures are sparse and spread through the patient’s three-dimensional images. We present and evaluate a delineation error detection system, which guides clinicians to slices of three-dimensional images that contain potentially clinically relevant segmentation failures. We co-designed the DEDS with clinicians and refined it based on an observational study, which allowed us to characterize clinicians’ navigation patterns and the use of information sources like AI uncertainty and patients’ dose distributions. We evaluated the DEDS’ potential to speed up the QA process through a simulation study with a retrospective cohort of patients. Results indicate that speed-ups are the most significant when equipping the DEDS with information sources indicative of clinical priority, which prevents unnecessary edits.
Visual inspection of the segmentation ensemble permits understanding the main trends and detecting anomalies that might indicate segmentation failures. Using a spaghetti plot to visualize all ensemble members is straightforward but prone to clutter. Contour boxplots prevent clutter and extra complexity by distilling essential ensemble information, which permits more efficient ensemble inspection. Nevertheless, they are time-consuming to compute, reducing their practical value. We present Inclusion Depth for contour ensembles. Inclusion Depth yields per ensemble member centrality scores that allow characterizing the distribution of segmentation ensembles in terms of properties like the median, trimmed mean, confidence bands, and outliers. Compared to previous contour depth notions, Inclusion Depth is significantly faster, making it more applicable in practice for time-critical contexts like QA in adaptive RT. We show how Inclusion Depth permits creating contour boxplots for ensembles with hundreds of segmentations in seconds.
It is not uncommon for distinct representative shapes to co-occur within a contour ensemble. With ensembles created by clinicians, for instance, different institutions, training sessions, or experience levels can lead to distinct shapes (i.e., modes of variation) for the same structure. When trained on these data, deep ensemble AIs would yield similarly multimodal ensembles. In quality assessment, being able to extract these representatives would pave the way for new ensemble-based interactive segmentation workflows. Applying traditional contour depth notions to these multi-modal ensembles collapses the existing variation modes and can lead to uninformative centrality scores. To address this issue, we present the first framework for multi-modal contour depth, which also includes notable runtime improvements for depth computation. When used with Inclusion Depth, multi-modal contour depth permits clustering the different modes of variation and determining cluster-dependent scores that appropriately characterize the data. Variation modes can be then independently analyzed using uni-modal depth machinery like contour boxplots. xiii
The global perspective of contour depth methods, which consider the entire volume, may be insufficient when parts of the contours are noisy or when the resolution of the ensemble is too large to process within a reasonable time. Correlation clustering methods provide a solution by partitioning the spatial domain of the ensemble into highly correlated regions that can be used to localize analyses. Existing correlation clustering algorithms do not scale well as the resolution of the ensemble increases. We introduce the Local-to-Global Correlation Clustering (LoGCC) method, which partitions the ensemble’s spatial domain into coarser primitives, representing areas of consistent ensemble member behavior. Unlike previous correlation clustering methods, the proposed LoGCC achieves significantly faster runtimes by leveraging the ensemble’s spatial structure and decoupling computations into local and global steps. Like with Inclusion Depth, these speed gains enable LoGCC to analyze large datasets in time-critical fields such as adaptive radiotherapy (RT).
Throughout this dissertation, our approach focused on designing modular, flexible analysis methods applicable across different tasks and domains. We demonstrate how the delineation error detection system, multi-modal Inclusion Depth, and Local-to-Global Correlation Clustering support quality assessment in RT and extend to fields like meteorology. We also speculate on their potential as foundational elements for more complex workflows. For example, extracted modes of variation, which indicate representative shapes in the ensemble, could be repurposed as an interactive segmentation tool. Alternatively, consistent regions detected by correlation clustering could be used as building blocks to enable localized contour analysis and editing.
We hope the proposed contour ensemble visual analysis methods inspire the development of more efficient analysis workflows that harness ensembles’ power in RT and beyond. ...
Radiotherapy (RT) is a widespread and effective technique to treat cancers by killing cancerous cells with rays of radiation. Building upon advances in image guidance and dose delivery technology like Proton Therapy, Adaptive RT promises more effective tumor decimation and a reduction of the incidence and severity of side effects. Unfortunately, the clinical implementation of adaptive workflows is challenging due to their resource-intensive nature. Therefore, their successful adoption lingers on overcoming several bottlenecks in the treatment planning process.
In this dissertation, we focus on methods used for the image segmentation or contouring step, which allows the localization of the anatomical structures required for dose optimization and evaluation. Until recently, clinicians had to manually delineate dozens of organs-at-risk and target volumes across hundreds of slices of the patient’s three-dimensional images. A process that is extremely time-consuming. The advent of deep learning-based artificial intelligence (AI) has changed the landscape: a modern auto-segmentation AI can produce segmentations for most of a patient’s anatomy in minutes.
Despite increasing automation in the segmentation process, it remains time and resource-intensive. Due to the segmentations’ criticality for the patient’s outcome and the errors the AI will commit, clinicians must perform a quality assessment of the AI’s outputs. Depending on the case’s complexity, the duration of the quality assessment process can negate the time gains auto-segmentation tools bring.
Deep ensemble AIs represent an advancement in medical image segmentation. Instead of providing a deterministic output, deep ensemble AIs produce a set of plausible candidates that aim to model inter-clinician annotation variability. Consensus segmentations obtained from ensembles tend to be more accurate and robust than the single-prediction deterministic counterpart. Nevertheless, by only using the consensus, a lot of potentially useful information is being discarded.
In this dissertation, we contribute to different phases of the segmentation quality assessment process. We characterize this process and introduce methods that leverage the raw outputs of deep ensemble AIs to support and speed up quality assessment tasks. The methods presented show new ways of analyzing and using ensembles in RT. Nevertheless, since these are relevant outside RT, we keep the presentation of the methods general and evaluate them in other application scenarios, such as the analysis of simulation ensembles or meteorological data.
Before fixing segmentation failures, clinicians must find them. This process can be time-consuming and fatiguing when failures are sparse and spread through the patient’s three-dimensional images. We present and evaluate a delineation error detection system, which guides clinicians to slices of three-dimensional images that contain potentially clinically relevant segmentation failures. We co-designed the DEDS with clinicians and refined it based on an observational study, which allowed us to characterize clinicians’ navigation patterns and the use of information sources like AI uncertainty and patients’ dose distributions. We evaluated the DEDS’ potential to speed up the QA process through a simulation study with a retrospective cohort of patients. Results indicate that speed-ups are the most significant when equipping the DEDS with information sources indicative of clinical priority, which prevents unnecessary edits.
Visual inspection of the segmentation ensemble permits understanding the main trends and detecting anomalies that might indicate segmentation failures. Using a spaghetti plot to visualize all ensemble members is straightforward but prone to clutter. Contour boxplots prevent clutter and extra complexity by distilling essential ensemble information, which permits more efficient ensemble inspection. Nevertheless, they are time-consuming to compute, reducing their practical value. We present Inclusion Depth for contour ensembles. Inclusion Depth yields per ensemble member centrality scores that allow characterizing the distribution of segmentation ensembles in terms of properties like the median, trimmed mean, confidence bands, and outliers. Compared to previous contour depth notions, Inclusion Depth is significantly faster, making it more applicable in practice for time-critical contexts like QA in adaptive RT. We show how Inclusion Depth permits creating contour boxplots for ensembles with hundreds of segmentations in seconds.
It is not uncommon for distinct representative shapes to co-occur within a contour ensemble. With ensembles created by clinicians, for instance, different institutions, training sessions, or experience levels can lead to distinct shapes (i.e., modes of variation) for the same structure. When trained on these data, deep ensemble AIs would yield similarly multimodal ensembles. In quality assessment, being able to extract these representatives would pave the way for new ensemble-based interactive segmentation workflows. Applying traditional contour depth notions to these multi-modal ensembles collapses the existing variation modes and can lead to uninformative centrality scores. To address this issue, we present the first framework for multi-modal contour depth, which also includes notable runtime improvements for depth computation. When used with Inclusion Depth, multi-modal contour depth permits clustering the different modes of variation and determining cluster-dependent scores that appropriately characterize the data. Variation modes can be then independently analyzed using uni-modal depth machinery like contour boxplots. xiii
The global perspective of contour depth methods, which consider the entire volume, may be insufficient when parts of the contours are noisy or when the resolution of the ensemble is too large to process within a reasonable time. Correlation clustering methods provide a solution by partitioning the spatial domain of the ensemble into highly correlated regions that can be used to localize analyses. Existing correlation clustering algorithms do not scale well as the resolution of the ensemble increases. We introduce the Local-to-Global Correlation Clustering (LoGCC) method, which partitions the ensemble’s spatial domain into coarser primitives, representing areas of consistent ensemble member behavior. Unlike previous correlation clustering methods, the proposed LoGCC achieves significantly faster runtimes by leveraging the ensemble’s spatial structure and decoupling computations into local and global steps. Like with Inclusion Depth, these speed gains enable LoGCC to analyze large datasets in time-critical fields such as adaptive radiotherapy (RT).
Throughout this dissertation, our approach focused on designing modular, flexible analysis methods applicable across different tasks and domains. We demonstrate how the delineation error detection system, multi-modal Inclusion Depth, and Local-to-Global Correlation Clustering support quality assessment in RT and extend to fields like meteorology. We also speculate on their potential as foundational elements for more complex workflows. For example, extracted modes of variation, which indicate representative shapes in the ensemble, could be repurposed as an interactive segmentation tool. Alternatively, consistent regions detected by correlation clustering could be used as building blocks to enable localized contour analysis and editing.
We hope the proposed contour ensemble visual analysis methods inspire the development of more efficient analysis workflows that harness ensembles’ power in RT and beyond.
In this dissertation, we focus on methods used for the image segmentation or contouring step, which allows the localization of the anatomical structures required for dose optimization and evaluation. Until recently, clinicians had to manually delineate dozens of organs-at-risk and target volumes across hundreds of slices of the patient’s three-dimensional images. A process that is extremely time-consuming. The advent of deep learning-based artificial intelligence (AI) has changed the landscape: a modern auto-segmentation AI can produce segmentations for most of a patient’s anatomy in minutes.
Despite increasing automation in the segmentation process, it remains time and resource-intensive. Due to the segmentations’ criticality for the patient’s outcome and the errors the AI will commit, clinicians must perform a quality assessment of the AI’s outputs. Depending on the case’s complexity, the duration of the quality assessment process can negate the time gains auto-segmentation tools bring.
Deep ensemble AIs represent an advancement in medical image segmentation. Instead of providing a deterministic output, deep ensemble AIs produce a set of plausible candidates that aim to model inter-clinician annotation variability. Consensus segmentations obtained from ensembles tend to be more accurate and robust than the single-prediction deterministic counterpart. Nevertheless, by only using the consensus, a lot of potentially useful information is being discarded.
In this dissertation, we contribute to different phases of the segmentation quality assessment process. We characterize this process and introduce methods that leverage the raw outputs of deep ensemble AIs to support and speed up quality assessment tasks. The methods presented show new ways of analyzing and using ensembles in RT. Nevertheless, since these are relevant outside RT, we keep the presentation of the methods general and evaluate them in other application scenarios, such as the analysis of simulation ensembles or meteorological data.
Before fixing segmentation failures, clinicians must find them. This process can be time-consuming and fatiguing when failures are sparse and spread through the patient’s three-dimensional images. We present and evaluate a delineation error detection system, which guides clinicians to slices of three-dimensional images that contain potentially clinically relevant segmentation failures. We co-designed the DEDS with clinicians and refined it based on an observational study, which allowed us to characterize clinicians’ navigation patterns and the use of information sources like AI uncertainty and patients’ dose distributions. We evaluated the DEDS’ potential to speed up the QA process through a simulation study with a retrospective cohort of patients. Results indicate that speed-ups are the most significant when equipping the DEDS with information sources indicative of clinical priority, which prevents unnecessary edits.
Visual inspection of the segmentation ensemble permits understanding the main trends and detecting anomalies that might indicate segmentation failures. Using a spaghetti plot to visualize all ensemble members is straightforward but prone to clutter. Contour boxplots prevent clutter and extra complexity by distilling essential ensemble information, which permits more efficient ensemble inspection. Nevertheless, they are time-consuming to compute, reducing their practical value. We present Inclusion Depth for contour ensembles. Inclusion Depth yields per ensemble member centrality scores that allow characterizing the distribution of segmentation ensembles in terms of properties like the median, trimmed mean, confidence bands, and outliers. Compared to previous contour depth notions, Inclusion Depth is significantly faster, making it more applicable in practice for time-critical contexts like QA in adaptive RT. We show how Inclusion Depth permits creating contour boxplots for ensembles with hundreds of segmentations in seconds.
It is not uncommon for distinct representative shapes to co-occur within a contour ensemble. With ensembles created by clinicians, for instance, different institutions, training sessions, or experience levels can lead to distinct shapes (i.e., modes of variation) for the same structure. When trained on these data, deep ensemble AIs would yield similarly multimodal ensembles. In quality assessment, being able to extract these representatives would pave the way for new ensemble-based interactive segmentation workflows. Applying traditional contour depth notions to these multi-modal ensembles collapses the existing variation modes and can lead to uninformative centrality scores. To address this issue, we present the first framework for multi-modal contour depth, which also includes notable runtime improvements for depth computation. When used with Inclusion Depth, multi-modal contour depth permits clustering the different modes of variation and determining cluster-dependent scores that appropriately characterize the data. Variation modes can be then independently analyzed using uni-modal depth machinery like contour boxplots. xiii
The global perspective of contour depth methods, which consider the entire volume, may be insufficient when parts of the contours are noisy or when the resolution of the ensemble is too large to process within a reasonable time. Correlation clustering methods provide a solution by partitioning the spatial domain of the ensemble into highly correlated regions that can be used to localize analyses. Existing correlation clustering algorithms do not scale well as the resolution of the ensemble increases. We introduce the Local-to-Global Correlation Clustering (LoGCC) method, which partitions the ensemble’s spatial domain into coarser primitives, representing areas of consistent ensemble member behavior. Unlike previous correlation clustering methods, the proposed LoGCC achieves significantly faster runtimes by leveraging the ensemble’s spatial structure and decoupling computations into local and global steps. Like with Inclusion Depth, these speed gains enable LoGCC to analyze large datasets in time-critical fields such as adaptive radiotherapy (RT).
Throughout this dissertation, our approach focused on designing modular, flexible analysis methods applicable across different tasks and domains. We demonstrate how the delineation error detection system, multi-modal Inclusion Depth, and Local-to-Global Correlation Clustering support quality assessment in RT and extend to fields like meteorology. We also speculate on their potential as foundational elements for more complex workflows. For example, extracted modes of variation, which indicate representative shapes in the ensemble, could be repurposed as an interactive segmentation tool. Alternatively, consistent regions detected by correlation clustering could be used as building blocks to enable localized contour analysis and editing.
We hope the proposed contour ensemble visual analysis methods inspire the development of more efficient analysis workflows that harness ensembles’ power in RT and beyond.
The expansion of human lifespan has increased the demands for healthcare services. Meeting these demands requires healthcare to evolve and be more efficient. Rapi dly advancing technology plays a crucial role in this evolution by automating tasks and supporting healthcare providers in their workflow. This requires a symbiotic collaboration between humans and digital systems. On the human side, healthcare providers understand the clinical context and are highly specialized in integrating information from several sources. Yet, they face limits in attention span and workload capacity. On the digital side, patient monitoring systems operate tirelessly and with precision, yet they lack the ability to interpret the clinical context of information. Consequently, effective patient care relies on each side playing their specific roles....
...
The expansion of human lifespan has increased the demands for healthcare services. Meeting these demands requires healthcare to evolve and be more efficient. Rapi dly advancing technology plays a crucial role in this evolution by automating tasks and supporting healthcare providers in their workflow. This requires a symbiotic collaboration between humans and digital systems. On the human side, healthcare providers understand the clinical context and are highly specialized in integrating information from several sources. Yet, they face limits in attention span and workload capacity. On the digital side, patient monitoring systems operate tirelessly and with precision, yet they lack the ability to interpret the clinical context of information. Consequently, effective patient care relies on each side playing their specific roles....
During the last decades, many organizations have gone through or attempted to go through digital transformations to create value for their customers, stakeholders and employees. Findings from a thorough literature research pointed out that organizations are forced to implement strategies to embrace the implications of the current digital revolution and improve performance by innovating with new technologies (Thrassou et al., 2021). It became clear that agility in an IT landscape is crucial for the success of digital transformations and survival in volatile market conditions (Châlons & Dufft, 2017). For Deloitte consultants the importance of agility in the IT landscape is clear, however, it is unclear how the agility of an IT landscape can best be evaluated.
The aim of this research was to create a framework for the evaluation of IT landscape agility, on behalf of the Technology Transformation and Acceleration (TT&A) team within Deloitte consulting. The goal, method and content of the evaluation framework was defined based on a combination of insights from seven expert interviews, observations during team and capability meetings, three different focus groups and extensive literature research on the topics of digital transformation, information system agility and information technology agility. Using a design science research methodology, the framework has been iterated multiple times by moving between what is relevant for Deloitte and what is known in the literature sphere. Based on the final framework for evaluation, multiple concepts were developed, of which the quick scan survey was selected by the Deloitte consultants as most desirable. This concept was developed during multiple evaluations and iterations together with employees from the University of Technology Delft.
This process resulted in the IT Agility Scan, a survey that can be used during client engagements, enabling consultants to efficiently and effectively assess what the agility of an IT landscape. The evaluation considers the complete breadth of an organization, diving into technical dimensions, non-technical dimensions, and system characteristics. Involving different employees and stakeholders across organizational boundaries. Being timed based on project timelines, the IT Agility Scan will allow consultants to not only define the current state, but also track the impact of change. It is expected that applying the IT Agility Scan will improve the ability Deloitte consultants to evaluate clients by providing a standardized method for generating valuable insights in gaps and opportunities for the agility of a client IT landscape. The IT Agility Scan will decrease the time needed gathering and analyzing data during the assessment stage of client engagements.
The IT Agility Scan is accompanied by a recommended implementation strategy, illustrating the onboarding and optimization process. By making TT&A product owner, the position of TT&A within Deloitte is strengthened over time. Involving subject matter experts in the optimization process, ensures and stimulates information sharing between experts, teams and clients. To support the engagement and development process this report provides Deloitte with a recommended strategy for the next five years, presented in a tactical roadmap towards a holistic and automated IT Agility Scan. Lastly this report highlights the limitations of this project and recommendations for further research on the topic of IT landscape agility.
...
The aim of this research was to create a framework for the evaluation of IT landscape agility, on behalf of the Technology Transformation and Acceleration (TT&A) team within Deloitte consulting. The goal, method and content of the evaluation framework was defined based on a combination of insights from seven expert interviews, observations during team and capability meetings, three different focus groups and extensive literature research on the topics of digital transformation, information system agility and information technology agility. Using a design science research methodology, the framework has been iterated multiple times by moving between what is relevant for Deloitte and what is known in the literature sphere. Based on the final framework for evaluation, multiple concepts were developed, of which the quick scan survey was selected by the Deloitte consultants as most desirable. This concept was developed during multiple evaluations and iterations together with employees from the University of Technology Delft.
This process resulted in the IT Agility Scan, a survey that can be used during client engagements, enabling consultants to efficiently and effectively assess what the agility of an IT landscape. The evaluation considers the complete breadth of an organization, diving into technical dimensions, non-technical dimensions, and system characteristics. Involving different employees and stakeholders across organizational boundaries. Being timed based on project timelines, the IT Agility Scan will allow consultants to not only define the current state, but also track the impact of change. It is expected that applying the IT Agility Scan will improve the ability Deloitte consultants to evaluate clients by providing a standardized method for generating valuable insights in gaps and opportunities for the agility of a client IT landscape. The IT Agility Scan will decrease the time needed gathering and analyzing data during the assessment stage of client engagements.
The IT Agility Scan is accompanied by a recommended implementation strategy, illustrating the onboarding and optimization process. By making TT&A product owner, the position of TT&A within Deloitte is strengthened over time. Involving subject matter experts in the optimization process, ensures and stimulates information sharing between experts, teams and clients. To support the engagement and development process this report provides Deloitte with a recommended strategy for the next five years, presented in a tactical roadmap towards a holistic and automated IT Agility Scan. Lastly this report highlights the limitations of this project and recommendations for further research on the topic of IT landscape agility.
...
During the last decades, many organizations have gone through or attempted to go through digital transformations to create value for their customers, stakeholders and employees. Findings from a thorough literature research pointed out that organizations are forced to implement strategies to embrace the implications of the current digital revolution and improve performance by innovating with new technologies (Thrassou et al., 2021). It became clear that agility in an IT landscape is crucial for the success of digital transformations and survival in volatile market conditions (Châlons & Dufft, 2017). For Deloitte consultants the importance of agility in the IT landscape is clear, however, it is unclear how the agility of an IT landscape can best be evaluated.
The aim of this research was to create a framework for the evaluation of IT landscape agility, on behalf of the Technology Transformation and Acceleration (TT&A) team within Deloitte consulting. The goal, method and content of the evaluation framework was defined based on a combination of insights from seven expert interviews, observations during team and capability meetings, three different focus groups and extensive literature research on the topics of digital transformation, information system agility and information technology agility. Using a design science research methodology, the framework has been iterated multiple times by moving between what is relevant for Deloitte and what is known in the literature sphere. Based on the final framework for evaluation, multiple concepts were developed, of which the quick scan survey was selected by the Deloitte consultants as most desirable. This concept was developed during multiple evaluations and iterations together with employees from the University of Technology Delft.
This process resulted in the IT Agility Scan, a survey that can be used during client engagements, enabling consultants to efficiently and effectively assess what the agility of an IT landscape. The evaluation considers the complete breadth of an organization, diving into technical dimensions, non-technical dimensions, and system characteristics. Involving different employees and stakeholders across organizational boundaries. Being timed based on project timelines, the IT Agility Scan will allow consultants to not only define the current state, but also track the impact of change. It is expected that applying the IT Agility Scan will improve the ability Deloitte consultants to evaluate clients by providing a standardized method for generating valuable insights in gaps and opportunities for the agility of a client IT landscape. The IT Agility Scan will decrease the time needed gathering and analyzing data during the assessment stage of client engagements.
The IT Agility Scan is accompanied by a recommended implementation strategy, illustrating the onboarding and optimization process. By making TT&A product owner, the position of TT&A within Deloitte is strengthened over time. Involving subject matter experts in the optimization process, ensures and stimulates information sharing between experts, teams and clients. To support the engagement and development process this report provides Deloitte with a recommended strategy for the next five years, presented in a tactical roadmap towards a holistic and automated IT Agility Scan. Lastly this report highlights the limitations of this project and recommendations for further research on the topic of IT landscape agility.
The aim of this research was to create a framework for the evaluation of IT landscape agility, on behalf of the Technology Transformation and Acceleration (TT&A) team within Deloitte consulting. The goal, method and content of the evaluation framework was defined based on a combination of insights from seven expert interviews, observations during team and capability meetings, three different focus groups and extensive literature research on the topics of digital transformation, information system agility and information technology agility. Using a design science research methodology, the framework has been iterated multiple times by moving between what is relevant for Deloitte and what is known in the literature sphere. Based on the final framework for evaluation, multiple concepts were developed, of which the quick scan survey was selected by the Deloitte consultants as most desirable. This concept was developed during multiple evaluations and iterations together with employees from the University of Technology Delft.
This process resulted in the IT Agility Scan, a survey that can be used during client engagements, enabling consultants to efficiently and effectively assess what the agility of an IT landscape. The evaluation considers the complete breadth of an organization, diving into technical dimensions, non-technical dimensions, and system characteristics. Involving different employees and stakeholders across organizational boundaries. Being timed based on project timelines, the IT Agility Scan will allow consultants to not only define the current state, but also track the impact of change. It is expected that applying the IT Agility Scan will improve the ability Deloitte consultants to evaluate clients by providing a standardized method for generating valuable insights in gaps and opportunities for the agility of a client IT landscape. The IT Agility Scan will decrease the time needed gathering and analyzing data during the assessment stage of client engagements.
The IT Agility Scan is accompanied by a recommended implementation strategy, illustrating the onboarding and optimization process. By making TT&A product owner, the position of TT&A within Deloitte is strengthened over time. Involving subject matter experts in the optimization process, ensures and stimulates information sharing between experts, teams and clients. To support the engagement and development process this report provides Deloitte with a recommended strategy for the next five years, presented in a tactical roadmap towards a holistic and automated IT Agility Scan. Lastly this report highlights the limitations of this project and recommendations for further research on the topic of IT landscape agility.
This master thesis described the embodiment design process of a new full-range directional hi-fi loudspeaker using a cardioid radiation pattern for Dutch & Dutch. Directional speakers minimize room boundary reflections to ensure the listener listens only to the direct sound waves coming from the speaker instead of also listening to the non-direct sound waves that come from reflections against the wall, floor, and ceiling. Dutch & Dutch current 8c-speaker implement a directional cardioid radiation pattern for frequencies above 100Hz. However, as the human hearing goes to as low as 20Hz, a new concept was developed to accommodate a full-range cardioid radiation pattern.
An analysis of the company and its portfolio, sound acoustics, the target group, and design language is presented. The analysis provided guidelines to start the embodiment design, which focuses on aesthetics, materialization, production, and assembly. Experts on these different topics were consulted to provide expertise-feedback and evaluation of the designed solutions. Multiple prototypes were made for evaluation and communication purposes. The final design proposition beholds an embodied concept, including acoustical concept, aesthetical design, and materialization of a full-range directional high-end hi-fi loudspeaker reflecting the values of Dutch & Dutch. ...
An analysis of the company and its portfolio, sound acoustics, the target group, and design language is presented. The analysis provided guidelines to start the embodiment design, which focuses on aesthetics, materialization, production, and assembly. Experts on these different topics were consulted to provide expertise-feedback and evaluation of the designed solutions. Multiple prototypes were made for evaluation and communication purposes. The final design proposition beholds an embodied concept, including acoustical concept, aesthetical design, and materialization of a full-range directional high-end hi-fi loudspeaker reflecting the values of Dutch & Dutch. ...
This master thesis described the embodiment design process of a new full-range directional hi-fi loudspeaker using a cardioid radiation pattern for Dutch & Dutch. Directional speakers minimize room boundary reflections to ensure the listener listens only to the direct sound waves coming from the speaker instead of also listening to the non-direct sound waves that come from reflections against the wall, floor, and ceiling. Dutch & Dutch current 8c-speaker implement a directional cardioid radiation pattern for frequencies above 100Hz. However, as the human hearing goes to as low as 20Hz, a new concept was developed to accommodate a full-range cardioid radiation pattern.
An analysis of the company and its portfolio, sound acoustics, the target group, and design language is presented. The analysis provided guidelines to start the embodiment design, which focuses on aesthetics, materialization, production, and assembly. Experts on these different topics were consulted to provide expertise-feedback and evaluation of the designed solutions. Multiple prototypes were made for evaluation and communication purposes. The final design proposition beholds an embodied concept, including acoustical concept, aesthetical design, and materialization of a full-range directional high-end hi-fi loudspeaker reflecting the values of Dutch & Dutch.
An analysis of the company and its portfolio, sound acoustics, the target group, and design language is presented. The analysis provided guidelines to start the embodiment design, which focuses on aesthetics, materialization, production, and assembly. Experts on these different topics were consulted to provide expertise-feedback and evaluation of the designed solutions. Multiple prototypes were made for evaluation and communication purposes. The final design proposition beholds an embodied concept, including acoustical concept, aesthetical design, and materialization of a full-range directional high-end hi-fi loudspeaker reflecting the values of Dutch & Dutch.
Concept development for an immersive audio mastering application
Designing a tool for object-based audio mixing
Immersive audio is an upcoming innovation within the world of professional audio reinforcement. The amount of possibilities of enhancing the listening experience is enlarged drastically by setting up arrays of speakers in front of and around the audience. Technological advancements make it possible to render recorded audio channels for an immersive audio system during live performances. These advancements rely on spatial audio rendering techniques, which will be described partly in this thesis. The elaboration on these techniques serves to clarify how the created audio effects are generated. Object-based audio (OBA) mixing allows the rendering of (pre)recorded audio channels for an immersive audio system. OBA mixing allows to be very close to the listening experience. In the project that is presented in this thesis, an in-depth look is taken at the use cases of immersive audio in the application of live performances. The users of immersive audio systems, who carry the responsibility of developing and executing the content that is reproduced over such systems, are determined and involved in defining a number of common use cases. Through the ability of defining specific requirements, these use cases lay the foundation for the development of a concept of a mixing tool for immersive audio systems. These requirements are supported by a respective definition of quality. Qualitative research is also done in terms of generating ideas. Furthermore, a look is taken at existing relevant software, with analogous functionalities. The research is bundled in a concept for a mixing tool, which is worked out in a visual mock-up. This mock-up is used in an evaluation study, to create an insight into the desirability of the concept. The concept is also laid parallel to its requirements that have been set for it. At last, elements of the concept, which are missing despite their relevance, are discussed. Further research is suggested for the development of these elements. Additionally, some use cases that arose during the project for the use of OBA mixing in the further future are presented.
...
Immersive audio is an upcoming innovation within the world of professional audio reinforcement. The amount of possibilities of enhancing the listening experience is enlarged drastically by setting up arrays of speakers in front of and around the audience. Technological advancements make it possible to render recorded audio channels for an immersive audio system during live performances. These advancements rely on spatial audio rendering techniques, which will be described partly in this thesis. The elaboration on these techniques serves to clarify how the created audio effects are generated. Object-based audio (OBA) mixing allows the rendering of (pre)recorded audio channels for an immersive audio system. OBA mixing allows to be very close to the listening experience. In the project that is presented in this thesis, an in-depth look is taken at the use cases of immersive audio in the application of live performances. The users of immersive audio systems, who carry the responsibility of developing and executing the content that is reproduced over such systems, are determined and involved in defining a number of common use cases. Through the ability of defining specific requirements, these use cases lay the foundation for the development of a concept of a mixing tool for immersive audio systems. These requirements are supported by a respective definition of quality. Qualitative research is also done in terms of generating ideas. Furthermore, a look is taken at existing relevant software, with analogous functionalities. The research is bundled in a concept for a mixing tool, which is worked out in a visual mock-up. This mock-up is used in an evaluation study, to create an insight into the desirability of the concept. The concept is also laid parallel to its requirements that have been set for it. At last, elements of the concept, which are missing despite their relevance, are discussed. Further research is suggested for the development of these elements. Additionally, some use cases that arose during the project for the use of OBA mixing in the further future are presented.
Patient empowerment via a smartwatch activity coach application
Let the patient gain back contral over their physical and mental health condition
A new healthcare domain is growing which is called eHealth. eHealth solutions are getting deployed in order to tackle the ongoing problem of an increase in patients with a chronic disease in The Netherlands. Medicine Men has developed a eHealth solution for patients with chronic diseases which is called the Emma Activity Coach. This application runs on a Fitbit smartwatch and is part of the Emma system which consists out of a smartphone application called Emma.6 and a desktop application called Emma dashboard. The aim of the Emma Activity Coach and the Emma system is to empower patients with a chronic disease in monitoring their activity in their own environment with the support of an informal caregiver and a healthcare professional in order to improve the quality of life of the patient. Project focus The focus of this project is on exploring the target group of the Emma Activity Coach and its needs in order to improve the Emma system. The user interaction and experience of the user are explored. Problems that come up are translated into design iterations which result in a proposed final design that fulfils the found needs and the set design vison. The ecosystem of the project consist out of three parties: • The patient • The informal caregiver • The healthcare professional Target group Literature research and interviews with several healthcare professionals showed that the use of the Emma Activity Coach would be beneficial for patients with chronic diseases such as COPD, Type 2 diabetes and Cardiovascular Disease (CVD). The target group is characterised by comorbidity; I.e. suffers from more than one chronic disease. This means that there is an overlap of disease symptoms of which activity, and therefore the use of the Emma Activity Coach is found to be beneficial. Target group challenge and needs The target group is explored via several patient interviews who suffer from COPD and/or Type 2 diabetes, three informal caregivers and three healthcare professionals. The main challenge for the target group is knowing their activity limits in order to gain back trust in their physical and mental condition. The target group has problems with getting enough activity or sometimes getting too much activity, which is often the case for patients with COPD. Their disease makes it hard for them to recognize their limits due to constant being short of breath. The challenge of the target group is translated into four needs; the need for autonomy, security, stimulation and support. Several design criteria are setup according to these needs and translated in the concluding design vision; to make the patient feel in control of their physical and mental health condition, by providing clear feedback on their activity. Design process Several user studies with COPD patients who had experience with the use of the Emma Activity Coach on a Fitbit smartwatch were performed. Interviews, observations and user tests revealed multiple design and comprehensibility problems with the current design of the Emma Activity Coach. These problems were tackled via three iteration rounds with the target group participants. The Emma dashboard of the healthcare professional, with the focus on physiotherapists, is also explored by interviewing 4 physiotherapists who have made use or still use the Emma dashboard in combination with the Emma Activity Coach. The iteration is presented as a starting point for further development. The evaluation with the target group shows that the proposed design of the Emma Activity Coach achieves the set design vision of this project.
...
A new healthcare domain is growing which is called eHealth. eHealth solutions are getting deployed in order to tackle the ongoing problem of an increase in patients with a chronic disease in The Netherlands. Medicine Men has developed a eHealth solution for patients with chronic diseases which is called the Emma Activity Coach. This application runs on a Fitbit smartwatch and is part of the Emma system which consists out of a smartphone application called Emma.6 and a desktop application called Emma dashboard. The aim of the Emma Activity Coach and the Emma system is to empower patients with a chronic disease in monitoring their activity in their own environment with the support of an informal caregiver and a healthcare professional in order to improve the quality of life of the patient. Project focus The focus of this project is on exploring the target group of the Emma Activity Coach and its needs in order to improve the Emma system. The user interaction and experience of the user are explored. Problems that come up are translated into design iterations which result in a proposed final design that fulfils the found needs and the set design vison. The ecosystem of the project consist out of three parties: • The patient • The informal caregiver • The healthcare professional Target group Literature research and interviews with several healthcare professionals showed that the use of the Emma Activity Coach would be beneficial for patients with chronic diseases such as COPD, Type 2 diabetes and Cardiovascular Disease (CVD). The target group is characterised by comorbidity; I.e. suffers from more than one chronic disease. This means that there is an overlap of disease symptoms of which activity, and therefore the use of the Emma Activity Coach is found to be beneficial. Target group challenge and needs The target group is explored via several patient interviews who suffer from COPD and/or Type 2 diabetes, three informal caregivers and three healthcare professionals. The main challenge for the target group is knowing their activity limits in order to gain back trust in their physical and mental condition. The target group has problems with getting enough activity or sometimes getting too much activity, which is often the case for patients with COPD. Their disease makes it hard for them to recognize their limits due to constant being short of breath. The challenge of the target group is translated into four needs; the need for autonomy, security, stimulation and support. Several design criteria are setup according to these needs and translated in the concluding design vision; to make the patient feel in control of their physical and mental health condition, by providing clear feedback on their activity. Design process Several user studies with COPD patients who had experience with the use of the Emma Activity Coach on a Fitbit smartwatch were performed. Interviews, observations and user tests revealed multiple design and comprehensibility problems with the current design of the Emma Activity Coach. These problems were tackled via three iteration rounds with the target group participants. The Emma dashboard of the healthcare professional, with the focus on physiotherapists, is also explored by interviewing 4 physiotherapists who have made use or still use the Emma dashboard in combination with the Emma Activity Coach. The iteration is presented as a starting point for further development. The evaluation with the target group shows that the proposed design of the Emma Activity Coach achieves the set design vision of this project.
Companies and content creators have mostly focused on the visual side of virtual reality. However, sound has the ability to improve the immersivity and perceived quality of a visual display and virtual simulation in general. In Social VR, it is especially important to be immersed. Social VR is a way of communicating with other people through VR by integrating participants in the same environment and enabling them to interact with each other. This graduation project focused on the design of an immersive sound experience during a Social VR session between two friends watching TV together. The purpose of this project was to evaluate current implementations of the sound experience in Social VR and to improve that experience through user testing. A literature research was done, Social VR demos were tested, questionnaires were made, and two prototypes were developed and tested. The first prototype was made for a basic and passive Social VR experience. The second one was a more advanced prototype in which two users could speak with each other, watch TV, interact with buttons and see each other’s avatar. Main findings: - Users expect the experience to feel like real life, with realistic visuals and sounds. The experience should be relaxing, immersing, and give the sense of being together. - Hearing yourself speak in VR through your headphones (sidetone) improves the experienc eand presence. - Users expect a home cinema experience for the TV, which can be done with a virtual 5.1 speaker setup without reverb. - Adding reverberation and spatialization to the voices improves co-presence and realism. - The reverberation of the room should be accurate as users could sense when it did not match their expectations. - The environment should be a reasonably large size for a living room, it should be open and detailed. It should also have a suitable background noise such as rain. ...
Companies and content creators have mostly focused on the visual side of virtual reality. However, sound has the ability to improve the immersivity and perceived quality of a visual display and virtual simulation in general. In Social VR, it is especially important to be immersed. Social VR is a way of communicating with other people through VR by integrating participants in the same environment and enabling them to interact with each other. This graduation project focused on the design of an immersive sound experience during a Social VR session between two friends watching TV together. The purpose of this project was to evaluate current implementations of the sound experience in Social VR and to improve that experience through user testing. A literature research was done, Social VR demos were tested, questionnaires were made, and two prototypes were developed and tested. The first prototype was made for a basic and passive Social VR experience. The second one was a more advanced prototype in which two users could speak with each other, watch TV, interact with buttons and see each other’s avatar. Main findings: - Users expect the experience to feel like real life, with realistic visuals and sounds. The experience should be relaxing, immersing, and give the sense of being together. - Hearing yourself speak in VR through your headphones (sidetone) improves the experienc eand presence. - Users expect a home cinema experience for the TV, which can be done with a virtual 5.1 speaker setup without reverb. - Adding reverberation and spatialization to the voices improves co-presence and realism. - The reverberation of the room should be accurate as users could sense when it did not match their expectations. - The environment should be a reasonably large size for a living room, it should be open and detailed. It should also have a suitable background noise such as rain.
Nowadays people become busier than ever before. Even though people still have 24 hours in each day, it feels as if no one has “free time” to do anything. Thus, busy people prefer using their limited time to clean the most important area indeed. It is how the idea “dust map” came. Users can create the map by themselves and the appliance will collect data to predict or explain the level of cleanliness at home. In this way, users are able to know which area should be cleaned the most and which area can be skipped out of their limited time. The project is to redesign the UI of Philips SpeedPro Max, based on this futuristic idea.
...
Nowadays people become busier than ever before. Even though people still have 24 hours in each day, it feels as if no one has “free time” to do anything. Thus, busy people prefer using their limited time to clean the most important area indeed. It is how the idea “dust map” came. Users can create the map by themselves and the appliance will collect data to predict or explain the level of cleanliness at home. In this way, users are able to know which area should be cleaned the most and which area can be skipped out of their limited time. The project is to redesign the UI of Philips SpeedPro Max, based on this futuristic idea.
The combination of centralised automatic bridge and lock (object) control and the increase in traffic has led to a shift in object operations. In the past years, several incidents and accidents have made RWS see the value of innovation when it comes to the control of objects. This master thesis will, therefore, start with defining which problems are relevant to operators and which solutions benefit them in object operations. To gather inside on the deeper knowledge, experiences and emotions of operators during the operation process, contextmapping was used. Contextmapping enhances the operator’s own understanding of experiences around safe object operations. To validate the contextmapping findings, 13 operators were interviewed using a combination of semi-structured interviews and the Scenes™ method. The results from the contextmapping sessions and interviews were visualised in an operator segmentation and an operator journey map. The segmentation allows for a better understanding of the target group, when to utilise their expertise during the innovation process and how they will react towards specific solutions. A design roadmap containing 5 horizons was created to provide RWS with an innovation strategy for the future of object control. In order to deliver an optimal service for both road and waterway traffic, while accommodating both RWS and the operator’s values the future vision regarding object operations will be: the future of object control will be an all-inclusive system to increase safety and traffic flow on water and land. The concepts proposed in these horizons all contribute to reaching this final vision. Furthermore, it is advised to keep involving the operators throughout every horizon using creative methods.
...
The combination of centralised automatic bridge and lock (object) control and the increase in traffic has led to a shift in object operations. In the past years, several incidents and accidents have made RWS see the value of innovation when it comes to the control of objects. This master thesis will, therefore, start with defining which problems are relevant to operators and which solutions benefit them in object operations. To gather inside on the deeper knowledge, experiences and emotions of operators during the operation process, contextmapping was used. Contextmapping enhances the operator’s own understanding of experiences around safe object operations. To validate the contextmapping findings, 13 operators were interviewed using a combination of semi-structured interviews and the Scenes™ method. The results from the contextmapping sessions and interviews were visualised in an operator segmentation and an operator journey map. The segmentation allows for a better understanding of the target group, when to utilise their expertise during the innovation process and how they will react towards specific solutions. A design roadmap containing 5 horizons was created to provide RWS with an innovation strategy for the future of object control. In order to deliver an optimal service for both road and waterway traffic, while accommodating both RWS and the operator’s values the future vision regarding object operations will be: the future of object control will be an all-inclusive system to increase safety and traffic flow on water and land. The concepts proposed in these horizons all contribute to reaching this final vision. Furthermore, it is advised to keep involving the operators throughout every horizon using creative methods.
This thesis focuses on the utilisation of additive manufacturing (AM) technologies for the production of accordions. The goal is to reduce the man hours required in the production and repair of the instrument. The client, Pigini Nederland, currently produces a small accordion for children and adults: the MiniMouse. This instrument is sold for 999 euros, which is a relatively low price for the amount of man hours invested. Pigini Nederland wants to lower the threshold of becoming familiar with the accordion by reducing the price of an entry-level instrument, similar to the MiniMouse in terms of functionality.
The sound of an accordion is produced by a reed: a piece of spring steel that vibrates when air flows past. To create a tone, the reed’s valve needs to be opened and an airflow needs to be created using the bellow. A mechanical structure of aluminium bars forms the connection between reed valve and button. A torsion spring keeps the valve in a closed position and creates resilience for the button.
The production of an accordion is a complex process consisting of mostly manual operations. Some of these operations are rather time and labour intensive, such as shaping the body and inner mechanics, and creating the bellow. The repair of an instrument can be an inefficient process: the complete disassembly of certain components is sometimes necessary to replace a single component.
Producing parts with complex geometry is one of the strengths of AM. This can lead to a reduction in tooling and inventory and part consolidation. This is an important driver for choosing AM as a means of production.
Fused Deposition Modeling (FDM) is chosen as the production technique for this project. FDM prints have good mechanical properties and require little post-processing. There is a wide range of materials available and the process and printer are relatively cheap. Pigini Nederland is interested in in-house production, which is realisable using an FDM printer.
The assignment focuses on the right hand side of an accordion, which has been fully designed and 3D printed. The fundament of the design is the instrument body. Multiple components are attached to it, resulting in a full-fledged instrument. For these attachments, non-printed connectors have been used as little as possible so that assembling the instrument is easy.
The mechanical structure consists of separate arms that are placed in the body using snap fits. A printed spring-like element is incorporated so that the arms of the structure bend when a button is pressed to open the reed valve. As the material loses its natural resilience during the expected 10 year product lifetime, a steel compression spring is added to regulate the button pushing force.
The buttons are attached to the mechanical structure using a snap fit. This makes it possible to quickly detach all buttons when repairing the instrument. In a conventional instrument, buttons are attached using glue and need to be broken off in such a scenario.
The reeds of a conventional accordion are attached using molten wax. Since this is labour-intensive during production and repair, the reeds in the printed instrument are clamped onto the body using a rubber gasket, nuts and bolts. The size of the reed sound chambers is determined by analysing sound samples and comparing them pairwise in a user test.
The project outcome provides an indication on how to use AM for accordion production. A printed proof of concept showcases that the instrument is fully functional, while minor design recommendations need to be addressed. An estimation of the material cost and labour during production is made, and a cost reduction of roughly 15% of the full instrument is established. This is a large step forward, as only the right hand side of the product has been redesigned. It is a clear indication that additive manufacturing can be a valuable tool in lowering the engagement threshold for future accordionists. ...
The sound of an accordion is produced by a reed: a piece of spring steel that vibrates when air flows past. To create a tone, the reed’s valve needs to be opened and an airflow needs to be created using the bellow. A mechanical structure of aluminium bars forms the connection between reed valve and button. A torsion spring keeps the valve in a closed position and creates resilience for the button.
The production of an accordion is a complex process consisting of mostly manual operations. Some of these operations are rather time and labour intensive, such as shaping the body and inner mechanics, and creating the bellow. The repair of an instrument can be an inefficient process: the complete disassembly of certain components is sometimes necessary to replace a single component.
Producing parts with complex geometry is one of the strengths of AM. This can lead to a reduction in tooling and inventory and part consolidation. This is an important driver for choosing AM as a means of production.
Fused Deposition Modeling (FDM) is chosen as the production technique for this project. FDM prints have good mechanical properties and require little post-processing. There is a wide range of materials available and the process and printer are relatively cheap. Pigini Nederland is interested in in-house production, which is realisable using an FDM printer.
The assignment focuses on the right hand side of an accordion, which has been fully designed and 3D printed. The fundament of the design is the instrument body. Multiple components are attached to it, resulting in a full-fledged instrument. For these attachments, non-printed connectors have been used as little as possible so that assembling the instrument is easy.
The mechanical structure consists of separate arms that are placed in the body using snap fits. A printed spring-like element is incorporated so that the arms of the structure bend when a button is pressed to open the reed valve. As the material loses its natural resilience during the expected 10 year product lifetime, a steel compression spring is added to regulate the button pushing force.
The buttons are attached to the mechanical structure using a snap fit. This makes it possible to quickly detach all buttons when repairing the instrument. In a conventional instrument, buttons are attached using glue and need to be broken off in such a scenario.
The reeds of a conventional accordion are attached using molten wax. Since this is labour-intensive during production and repair, the reeds in the printed instrument are clamped onto the body using a rubber gasket, nuts and bolts. The size of the reed sound chambers is determined by analysing sound samples and comparing them pairwise in a user test.
The project outcome provides an indication on how to use AM for accordion production. A printed proof of concept showcases that the instrument is fully functional, while minor design recommendations need to be addressed. An estimation of the material cost and labour during production is made, and a cost reduction of roughly 15% of the full instrument is established. This is a large step forward, as only the right hand side of the product has been redesigned. It is a clear indication that additive manufacturing can be a valuable tool in lowering the engagement threshold for future accordionists. ...
This thesis focuses on the utilisation of additive manufacturing (AM) technologies for the production of accordions. The goal is to reduce the man hours required in the production and repair of the instrument. The client, Pigini Nederland, currently produces a small accordion for children and adults: the MiniMouse. This instrument is sold for 999 euros, which is a relatively low price for the amount of man hours invested. Pigini Nederland wants to lower the threshold of becoming familiar with the accordion by reducing the price of an entry-level instrument, similar to the MiniMouse in terms of functionality.
The sound of an accordion is produced by a reed: a piece of spring steel that vibrates when air flows past. To create a tone, the reed’s valve needs to be opened and an airflow needs to be created using the bellow. A mechanical structure of aluminium bars forms the connection between reed valve and button. A torsion spring keeps the valve in a closed position and creates resilience for the button.
The production of an accordion is a complex process consisting of mostly manual operations. Some of these operations are rather time and labour intensive, such as shaping the body and inner mechanics, and creating the bellow. The repair of an instrument can be an inefficient process: the complete disassembly of certain components is sometimes necessary to replace a single component.
Producing parts with complex geometry is one of the strengths of AM. This can lead to a reduction in tooling and inventory and part consolidation. This is an important driver for choosing AM as a means of production.
Fused Deposition Modeling (FDM) is chosen as the production technique for this project. FDM prints have good mechanical properties and require little post-processing. There is a wide range of materials available and the process and printer are relatively cheap. Pigini Nederland is interested in in-house production, which is realisable using an FDM printer.
The assignment focuses on the right hand side of an accordion, which has been fully designed and 3D printed. The fundament of the design is the instrument body. Multiple components are attached to it, resulting in a full-fledged instrument. For these attachments, non-printed connectors have been used as little as possible so that assembling the instrument is easy.
The mechanical structure consists of separate arms that are placed in the body using snap fits. A printed spring-like element is incorporated so that the arms of the structure bend when a button is pressed to open the reed valve. As the material loses its natural resilience during the expected 10 year product lifetime, a steel compression spring is added to regulate the button pushing force.
The buttons are attached to the mechanical structure using a snap fit. This makes it possible to quickly detach all buttons when repairing the instrument. In a conventional instrument, buttons are attached using glue and need to be broken off in such a scenario.
The reeds of a conventional accordion are attached using molten wax. Since this is labour-intensive during production and repair, the reeds in the printed instrument are clamped onto the body using a rubber gasket, nuts and bolts. The size of the reed sound chambers is determined by analysing sound samples and comparing them pairwise in a user test.
The project outcome provides an indication on how to use AM for accordion production. A printed proof of concept showcases that the instrument is fully functional, while minor design recommendations need to be addressed. An estimation of the material cost and labour during production is made, and a cost reduction of roughly 15% of the full instrument is established. This is a large step forward, as only the right hand side of the product has been redesigned. It is a clear indication that additive manufacturing can be a valuable tool in lowering the engagement threshold for future accordionists.
The sound of an accordion is produced by a reed: a piece of spring steel that vibrates when air flows past. To create a tone, the reed’s valve needs to be opened and an airflow needs to be created using the bellow. A mechanical structure of aluminium bars forms the connection between reed valve and button. A torsion spring keeps the valve in a closed position and creates resilience for the button.
The production of an accordion is a complex process consisting of mostly manual operations. Some of these operations are rather time and labour intensive, such as shaping the body and inner mechanics, and creating the bellow. The repair of an instrument can be an inefficient process: the complete disassembly of certain components is sometimes necessary to replace a single component.
Producing parts with complex geometry is one of the strengths of AM. This can lead to a reduction in tooling and inventory and part consolidation. This is an important driver for choosing AM as a means of production.
Fused Deposition Modeling (FDM) is chosen as the production technique for this project. FDM prints have good mechanical properties and require little post-processing. There is a wide range of materials available and the process and printer are relatively cheap. Pigini Nederland is interested in in-house production, which is realisable using an FDM printer.
The assignment focuses on the right hand side of an accordion, which has been fully designed and 3D printed. The fundament of the design is the instrument body. Multiple components are attached to it, resulting in a full-fledged instrument. For these attachments, non-printed connectors have been used as little as possible so that assembling the instrument is easy.
The mechanical structure consists of separate arms that are placed in the body using snap fits. A printed spring-like element is incorporated so that the arms of the structure bend when a button is pressed to open the reed valve. As the material loses its natural resilience during the expected 10 year product lifetime, a steel compression spring is added to regulate the button pushing force.
The buttons are attached to the mechanical structure using a snap fit. This makes it possible to quickly detach all buttons when repairing the instrument. In a conventional instrument, buttons are attached using glue and need to be broken off in such a scenario.
The reeds of a conventional accordion are attached using molten wax. Since this is labour-intensive during production and repair, the reeds in the printed instrument are clamped onto the body using a rubber gasket, nuts and bolts. The size of the reed sound chambers is determined by analysing sound samples and comparing them pairwise in a user test.
The project outcome provides an indication on how to use AM for accordion production. A printed proof of concept showcases that the instrument is fully functional, while minor design recommendations need to be addressed. An estimation of the material cost and labour during production is made, and a cost reduction of roughly 15% of the full instrument is established. This is a large step forward, as only the right hand side of the product has been redesigned. It is a clear indication that additive manufacturing can be a valuable tool in lowering the engagement threshold for future accordionists.
This report is the final deliverable for the master track ‘Design for Interaction’ of the faculty of Industrial Design Engineering at the Technical University of Delft. It describes the process and results of the graduation project ‘Assuring the occurrence of intended learning situations in a nautical simulator’ that is executed in cooperation with VSTEP, a Dutch company that develops simulators and visual training software. VSTEP learned from its customers that Nautis is suffering from usability issues. It is initially not described where these issues are located. The project therefore starts with desk research, analysing Nautis’ current system. Additional expert interviews are conducted and literature research is executed. Observations and semi-structured interviews were conducted with instructors, the endusers, at five organisations that use Nautis. This has demonstrated that instructors are have a relatively small fraction of time available to develop scenarios. A scenario implies a virtual environment in which target vessels are following pre-specified routes, in order to mimic nautical traffic and thereupon form learning situations for the trainee. A significant amount of the little available time goes into testing and subsequently modifying. Testing usually occurs by the instructor entering one of the bridges connected to the simulator and sailing through his created scenario. Testing suggests verifying the behaviour of target vessels along their respective routes in order to prevent collisions with others, while at the same time checking if learning situations are encountered by a trainee vessel as intended. Modifications are made to the scenario when it does not come out as desired, and extra learning situations are added to accommodate different trainee performances, i.e., their pace through the scenario. This would allow for a better comparison of trainee performances. After modifying a scenario, it is again tested. Nautis offers an inefficient way of developing scenarios, which leaves instructors doubting if a trainee will reach the training goals properly. Therefore it is decided to focus on creating a solution for instructors which them to efficiently design learning situations in scenarios that allow for comparing trainee performance in order to assure that learning goals are reached. A design is proposal consisting of three components; an event creator, a trigger system, and a route generating algorithm. Each component supports the user in efficiently creating learning situations as intended. The first two components have been designed through an iterative process, suggesting sketching, prototyping, and user testing. Seven prototypes were made to validate concepts and test interactions with 13 unique participants. Some participants have tested multiple prototypes. Three instructors have taken part in the user tests. The algorithm is evaluated on its feasibility with an AI expert and desirability with an instructor. The route creator enables the user to efficiently draw out learning situations. A timeline is used to give the instructor quick access to any given timeframe in his scenario. When drawing a route for a target vessel, the currently displayed timeframe moves forward concordantly to the time it takes the particular vessel to travel the distance between the previous and the new waypoint. All target vessels in the scenario are displayed at their respective locations associated to the current timeframe. This offers the instructors direct insight whether a collision occurs, and can as such be immediately avoided. The trigger system allows the instructor to connect specific locations for target vessels in a learning situation to a manually drawn trigger area. The system assures that the target vessels sail their routes following the specified locations when the trainee enters the area with his vessel. This way the learning situation happens as intended by the instructor, independent of the trainee’s performance up until the trigger area. The route generating algorithm is used to save time on scenario development. Instructors are enabled to focus on drawing relevant parts of routes for target vessels that form learning situations. Parts that are deemed irrelevant, i.e., prior to or after a learning situation, are generated to lead the target vessels out of the trainees view. It is calculated that through the design proposal instructors can save more than 80% of their time on the development of a single scenario, ergo more time can be spend on the quality and the variety of the scenarios. Furthermore provides the assurance of the occurrence of learning situations them more guidance during the actual training sessions, and offers better comparison of trainee performances. To conclude, it can thus be stated that the design proposal would support an improvement in educational use of Nautis.
...
This report is the final deliverable for the master track ‘Design for Interaction’ of the faculty of Industrial Design Engineering at the Technical University of Delft. It describes the process and results of the graduation project ‘Assuring the occurrence of intended learning situations in a nautical simulator’ that is executed in cooperation with VSTEP, a Dutch company that develops simulators and visual training software. VSTEP learned from its customers that Nautis is suffering from usability issues. It is initially not described where these issues are located. The project therefore starts with desk research, analysing Nautis’ current system. Additional expert interviews are conducted and literature research is executed. Observations and semi-structured interviews were conducted with instructors, the endusers, at five organisations that use Nautis. This has demonstrated that instructors are have a relatively small fraction of time available to develop scenarios. A scenario implies a virtual environment in which target vessels are following pre-specified routes, in order to mimic nautical traffic and thereupon form learning situations for the trainee. A significant amount of the little available time goes into testing and subsequently modifying. Testing usually occurs by the instructor entering one of the bridges connected to the simulator and sailing through his created scenario. Testing suggests verifying the behaviour of target vessels along their respective routes in order to prevent collisions with others, while at the same time checking if learning situations are encountered by a trainee vessel as intended. Modifications are made to the scenario when it does not come out as desired, and extra learning situations are added to accommodate different trainee performances, i.e., their pace through the scenario. This would allow for a better comparison of trainee performances. After modifying a scenario, it is again tested. Nautis offers an inefficient way of developing scenarios, which leaves instructors doubting if a trainee will reach the training goals properly. Therefore it is decided to focus on creating a solution for instructors which them to efficiently design learning situations in scenarios that allow for comparing trainee performance in order to assure that learning goals are reached. A design is proposal consisting of three components; an event creator, a trigger system, and a route generating algorithm. Each component supports the user in efficiently creating learning situations as intended. The first two components have been designed through an iterative process, suggesting sketching, prototyping, and user testing. Seven prototypes were made to validate concepts and test interactions with 13 unique participants. Some participants have tested multiple prototypes. Three instructors have taken part in the user tests. The algorithm is evaluated on its feasibility with an AI expert and desirability with an instructor. The route creator enables the user to efficiently draw out learning situations. A timeline is used to give the instructor quick access to any given timeframe in his scenario. When drawing a route for a target vessel, the currently displayed timeframe moves forward concordantly to the time it takes the particular vessel to travel the distance between the previous and the new waypoint. All target vessels in the scenario are displayed at their respective locations associated to the current timeframe. This offers the instructors direct insight whether a collision occurs, and can as such be immediately avoided. The trigger system allows the instructor to connect specific locations for target vessels in a learning situation to a manually drawn trigger area. The system assures that the target vessels sail their routes following the specified locations when the trainee enters the area with his vessel. This way the learning situation happens as intended by the instructor, independent of the trainee’s performance up until the trigger area. The route generating algorithm is used to save time on scenario development. Instructors are enabled to focus on drawing relevant parts of routes for target vessels that form learning situations. Parts that are deemed irrelevant, i.e., prior to or after a learning situation, are generated to lead the target vessels out of the trainees view. It is calculated that through the design proposal instructors can save more than 80% of their time on the development of a single scenario, ergo more time can be spend on the quality and the variety of the scenarios. Furthermore provides the assurance of the occurrence of learning situations them more guidance during the actual training sessions, and offers better comparison of trainee performances. To conclude, it can thus be stated that the design proposal would support an improvement in educational use of Nautis.
Schiphol Security Scanner
Restoring the balance between passenger, agent and Scanner
Using Schiphol’s Security Scanner is confusing for a lot of passengers. Therefore, security agents currently have to fully take on the task of guiding the passenger through the Scanner. This repetitive task in the high volume environment of Schiphol security leads to physical and mental exhaustion which results in frustration, and is projected onto passengers.
The responsive animations concept guides passengers in taking the correct posture inside the Security Scanner. Real time skeletal tracking is done, whereafter the appropriate instructions and corrections are displayed. This concept aims to replace the instructive tasks of the agents to lighten their workload, and to defuse the tensions between passenger and agent.
Prototyping tests were executed in a live security operation to assess the effectiveness of the design and to record the agents’ experiences. Agents found the concept to work de-escalating because it acted as a mediator between passenger and agents. Moreover, they noticed a significant decrease in repetitive workload improving their overall mood and resilience. ...
The responsive animations concept guides passengers in taking the correct posture inside the Security Scanner. Real time skeletal tracking is done, whereafter the appropriate instructions and corrections are displayed. This concept aims to replace the instructive tasks of the agents to lighten their workload, and to defuse the tensions between passenger and agent.
Prototyping tests were executed in a live security operation to assess the effectiveness of the design and to record the agents’ experiences. Agents found the concept to work de-escalating because it acted as a mediator between passenger and agents. Moreover, they noticed a significant decrease in repetitive workload improving their overall mood and resilience. ...
Using Schiphol’s Security Scanner is confusing for a lot of passengers. Therefore, security agents currently have to fully take on the task of guiding the passenger through the Scanner. This repetitive task in the high volume environment of Schiphol security leads to physical and mental exhaustion which results in frustration, and is projected onto passengers.
The responsive animations concept guides passengers in taking the correct posture inside the Security Scanner. Real time skeletal tracking is done, whereafter the appropriate instructions and corrections are displayed. This concept aims to replace the instructive tasks of the agents to lighten their workload, and to defuse the tensions between passenger and agent.
Prototyping tests were executed in a live security operation to assess the effectiveness of the design and to record the agents’ experiences. Agents found the concept to work de-escalating because it acted as a mediator between passenger and agents. Moreover, they noticed a significant decrease in repetitive workload improving their overall mood and resilience.
The responsive animations concept guides passengers in taking the correct posture inside the Security Scanner. Real time skeletal tracking is done, whereafter the appropriate instructions and corrections are displayed. This concept aims to replace the instructive tasks of the agents to lighten their workload, and to defuse the tensions between passenger and agent.
Prototyping tests were executed in a live security operation to assess the effectiveness of the design and to record the agents’ experiences. Agents found the concept to work de-escalating because it acted as a mediator between passenger and agents. Moreover, they noticed a significant decrease in repetitive workload improving their overall mood and resilience.
Using Schiphol’s Security Scanner is confusing for a lot of passengers. Therefore, security agents currently have to fully take on the task of guiding the passenger through the Scanner. This repetitive task in the high volume environment of Schiphol security leads to physical and mental exhaustion and results in frustration, which is projected onto passengers.
The responsive animations concept guides passengers in taking the correct posture inside the security scanner. Real time skeletal tracking is done, whereafter the appropriate instructions and corrections are displayed. This concept aims to replace the instructive tasks of the agents to lighten their workload, and to defuse the tensions between passenger and agent.
Prototyping tests were executed in a live security operation to assess the effectiveness of the design and to record the agents’ experiences. Agents found the concept to work de-escalating because it acted as a mediator between passenger and agents. Moreover, they noticed a significant decrease in repetitive workload improving their overall mood and resilience. ...
The responsive animations concept guides passengers in taking the correct posture inside the security scanner. Real time skeletal tracking is done, whereafter the appropriate instructions and corrections are displayed. This concept aims to replace the instructive tasks of the agents to lighten their workload, and to defuse the tensions between passenger and agent.
Prototyping tests were executed in a live security operation to assess the effectiveness of the design and to record the agents’ experiences. Agents found the concept to work de-escalating because it acted as a mediator between passenger and agents. Moreover, they noticed a significant decrease in repetitive workload improving their overall mood and resilience. ...
Using Schiphol’s Security Scanner is confusing for a lot of passengers. Therefore, security agents currently have to fully take on the task of guiding the passenger through the Scanner. This repetitive task in the high volume environment of Schiphol security leads to physical and mental exhaustion and results in frustration, which is projected onto passengers.
The responsive animations concept guides passengers in taking the correct posture inside the security scanner. Real time skeletal tracking is done, whereafter the appropriate instructions and corrections are displayed. This concept aims to replace the instructive tasks of the agents to lighten their workload, and to defuse the tensions between passenger and agent.
Prototyping tests were executed in a live security operation to assess the effectiveness of the design and to record the agents’ experiences. Agents found the concept to work de-escalating because it acted as a mediator between passenger and agents. Moreover, they noticed a significant decrease in repetitive workload improving their overall mood and resilience.
The responsive animations concept guides passengers in taking the correct posture inside the security scanner. Real time skeletal tracking is done, whereafter the appropriate instructions and corrections are displayed. This concept aims to replace the instructive tasks of the agents to lighten their workload, and to defuse the tensions between passenger and agent.
Prototyping tests were executed in a live security operation to assess the effectiveness of the design and to record the agents’ experiences. Agents found the concept to work de-escalating because it acted as a mediator between passenger and agents. Moreover, they noticed a significant decrease in repetitive workload improving their overall mood and resilience.
Under Pressure
Explorations on the dynamics of prioritization in dual-task driving
Monitoring radio messages while driving is an omnipresent dual-task combination in police work, but it is also one that is considered unsafe for regular drivers. Whereas regular drivers are expected to fully prioritize the driving task, police officers typically do not have the option to stop their car to attend important incoming messages, nor can they afford an uninformed arrival at the scene. A novel method for the visualization of observational data shows that police work is highly fragmented, and suggests that frequent reports on work overload are related to dual- task involvement in this fragmented workflow. Therefore, a series of experimental studies have been conducted to understand the mechanics that underlie and result from task prioritization in a dynamic complex socio-technical system, such as the police context. Methodological implications are presented for the interpretation of tradeoffs between task performance and mental effort as function of task prioritization. Furthermore, practical implications are presented for the development of information technology in police vehicles. Finally, recommendations for future research include the validation of an integrated model on coping strategies, task prioritization, and dual-task switching.
...
Monitoring radio messages while driving is an omnipresent dual-task combination in police work, but it is also one that is considered unsafe for regular drivers. Whereas regular drivers are expected to fully prioritize the driving task, police officers typically do not have the option to stop their car to attend important incoming messages, nor can they afford an uninformed arrival at the scene. A novel method for the visualization of observational data shows that police work is highly fragmented, and suggests that frequent reports on work overload are related to dual- task involvement in this fragmented workflow. Therefore, a series of experimental studies have been conducted to understand the mechanics that underlie and result from task prioritization in a dynamic complex socio-technical system, such as the police context. Methodological implications are presented for the interpretation of tradeoffs between task performance and mental effort as function of task prioritization. Furthermore, practical implications are presented for the development of information technology in police vehicles. Finally, recommendations for future research include the validation of an integrated model on coping strategies, task prioritization, and dual-task switching.