ABSTRACT
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Embedded systems are a class of computer systems that are typically characterized by a tight interaction with the physical environment. Various methodologies have been adopted for the development of such systems, ranging from traditional... more
Embedded systems are a class of computer systems that are typically characterized by a tight interaction with the physical environment. Various methodologies have been adopted for the development of such systems, ranging from traditional waterfall to modern agile techniques. One of the agile techniques that has recently attracted increasing attention is Behavior-Driven Development (BDD). BDD promotes the engagement of all stakeholders in every development iteration to minimize the misunderstanding between technical and non-technical stakeholders and, consequently, to speed up the development process and lower the costs. In this paper, we investigate the application of BDD to the development of embedded systems, especially focusing on the testing of timing requirements for real-time embedded software. In particular, we extend BDD with time-related concepts and propose an approach to generate test code for the verification of timing behavior of real-time embedded systems. Our approach offers more automation for the development of test code compared to existing BDD tools, thus minimizing the risk of timing faults and reducing development costs and time-to-market.
Research Interests:
Simulations are widely used in the engineering workflow of complex mechatronic embedded systems in various domains, such as healthcare, railway, automotive and aerospace, for analyzing, testing and validating purposes. This paper focuses... more
Simulations are widely used in the engineering workflow of complex mechatronic embedded systems in various domains, such as healthcare, railway, automotive and aerospace, for analyzing, testing and validating purposes. This paper focuses on the development and test of the control software of complex mechatronic embedded systems from the perspective of software interfaces (e.g., driver APIs) and presents a systematic approach for testing the control software during the various stages of an engineering process. Since we assume that the physical (hardware) components of an under-control plant could be replaced with simulation models, various kinds of in-the-loop simulations, ranging from MiL to HiL, can be consequently acquired. Additionally, we present a mathematical model of MESes required to formally describe the approach and also a healthcare case study to which our approach was applied.
Research Interests:
Sport classification is a crucial step for content analysis in a sport stream monitoring system. Training a reliable sport classifier can be a challenging task when the data is limited in amount and highly imbalanced. In this paper, we... more
Sport classification is a crucial step for content analysis in a sport stream monitoring system. Training a reliable sport classifier can be a challenging task when the data is limited in amount and highly imbalanced. In this paper, we introduce a supervised two-stage transfer learning (Two-Stage-TL) method to solve the data shortage problem. It can progressively transfer features from a source domain to the target domain using a properly selected bridge domain. For the class imbalance issue, we compare several existing methods and demonstrate that the log-smoothing class weight is the most applicable way for this specific problem. Extensive experiments are conducted using ResNet50, VGG16, and Inception-ResNet-v2. The results show that Two-Stage-TL outperforms classical One-Stage-TL and achieves the best performance using log-smoothing class weight. The in-depth analysis is useful for researchers and developers in solving similar problems.
Research Interests:
ABSTRACT
Research Interests:
Research Interests:
Applications employing data classification such as smart lighting that involve human factors such as perception lead to non-deterministic input-output relationships where more than one output may be acceptable for a given input. For these... more
Applications employing data classification such as smart lighting that involve human factors such as perception lead to non-deterministic input-output relationships where more than one output may be acceptable for a given input. For these so called non-deterministic multiple output classification (nDMOC) problems, the relationship between the input and output may change over time making it difficult for the machine learning (ML) algorithms in a batch setting to make predictions for a given context. In this paper, we describe the nature of nDMOC problems and discuss the Relevance Score (RS) that is suitable in this context as a performance metric. RS determines the extent by which a predicted output is relevant to the user's context and behaviors, taking into account the inconsistencies that come with human (perception) factors. We tailor the RS metric so that it can be used to evaluate ML algorithms in an online setting at run-time. We assess the performance of a number of ML algorithms, using a smart lighting dataset with non-deterministic one-to-many input-output relationships. The results indicate that using RS instead of classification accuracy (CA) is suitable to analyze the performance of conventional ML algorithms applied to the category of nDMOC problems. Instance-based online ML gives the best RS performance. An interesting finding is that the RS keeps increasing with increasing number of samples, even after the CA performance converges.
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
... Diana Albu and Johan Lukkien and Richard Verhoeven Department of Mathematics and Computer Science, Eindhoven University of Technology POBox 513, 5600 MB Eindhoven, The Netherlands Email: http://www.win.tue.nl/san ... [16] I.... more
... Diana Albu and Johan Lukkien and Richard Verhoeven Department of Mathematics and Computer Science, Eindhoven University of Technology POBox 513, 5600 MB Eindhoven, The Netherlands Email: http://www.win.tue.nl/san ... [16] I. Provaznik, J. Kozumplik, J. Bardonova, M ...
Research Interests:
ABSTRACT
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
This paper proposes an alternative approach for evidence based trust computation where the relationship between evidence and trust is learned using an artificial neuron, making it possible to automatically adapt trust computation to... more
This paper proposes an alternative approach for evidence based trust computation where the relationship between evidence and trust is learned using an artificial neuron, making it possible to automatically adapt trust computation to different use cases. Computational trust aims to quantify trust based on ever increasing evidence on observations. In the literature a trust value is seen as a posterior subjective probability, computed using Bayesian inference on evidence, a prior and a weight of the prior. This provides a fixed mapping between evidence and trust, which may not be suitable for every case study, e.g. when positive and negative evidences are not equally important. The proposed solution is also a first step towards our future work to replace complex and case-specific trust fusion operators proposed in the literature with a generic case-independent artificial neural network solution. Our experiments on example cases of coin toss prediction and occupancy detection show that for sufficiently large data sets, i.e. given sufficient evidence based on a history of observations, the proposed learning approach yields comparable results and in some cases beats the existing approach.
Research Interests: Information Systems, Computer Science, Artificial Intelligence, Machine Learning, Inference, and 7 moreComputational Trust, Computation, Approximate Bayesian Computation, Bayesian Network, Computer Networks and Communications, Information Systems Development And Management, and Artificial Neural Network
Complex embedded systems, such as robotics, automotive and high-tech manufacturing, are hard to maintain due to their complex nature. To advance our understanding of the software engineering practice for complex embedded systems, we... more
Complex embedded systems, such as robotics, automotive and high-tech manufacturing, are hard to maintain due to their complex nature. To advance our understanding of the software engineering practice for complex embedded systems, we conducted a series of empirical studies at ASML, a leading manufacturer of lithography machines for semi-conductor industry. We started with an interview study exploring how developers use execution logs, essential artifacts that capture the runtime behavior of software systems. The empirical insights obtained from this study led us to explore subtopics about model inference from logs, modeling practice and log comparison. Motivated by the observation that developers often manually sketch behavioral models based on logs, we propose a model inference technique that can extract models by combining log analysis, and analysis of a running system under stimuli. As observed in this model inference study, the transition from code to models requires developers to work with a hybrid system which consists of handwritten code and models. We then study modeling practices and the roles of model in such hybrid systems. Particularly, we study why developers violate modeling guidelines, providing implications for researchers and tool builders to support developers in modeling complex embedded systems. Another interesting observation from the interview study is that developers face challenges in comparing multiple logs generated from such systems. We therefore conduct a literature study to provide an overview of the existing techniques and identify the limitations of the existing techniques. In this project, we study logs and models in complex embedded systems, providing tool builders, researchers and practitioners with implications to facilitate log analysis, model inference, modeling practice and log comparison.
Research Interests:
... of the proposed RPDM framework is one of the first open implementations based ... http://www.openmobilealliance.org/tech/affiliates/syncml/sync ml_dm_represent_v11_20020215. pdf ... Universal Manager seamless management of enterprise... more
... of the proposed RPDM framework is one of the first open implementations based ... http://www.openmobilealliance.org/tech/affiliates/syncml/sync ml_dm_represent_v11_20020215. pdf ... Universal Manager seamless management of enterprise mobile and non-mobile devices, ...
Research Interests:
Abstract The target of a smart lighting system is to control light sources in an environment (eg home, office) adaptively according to user contexts and preferences. Literature work in this area focuses on traditional light sources such... more
Abstract The target of a smart lighting system is to control light sources in an environment (eg home, office) adaptively according to user contexts and preferences. Literature work in this area focuses on traditional light sources such as incandescent and fluorescent lights, whereas this paper takes a step towards adopting LED luminaries. A novel illumination model for distributed LED luminary control is presented. A prototype system is designed and implemented using several LED luminaries and light sensors. Experiments carried out on ...
Research Interests:
Component-based software construction is of much interest in software engineering due to its promise of black-box composition and independent deployment. Because of this black-box nature, users of components must be concerned with their... more
Component-based software construction is of much interest in software engineering due to its promise of black-box composition and independent deployment. Because of this black-box nature, users of components must be concerned with their trustworthiness, in particular when third parties are involved. Component certification is used to improve trust, usually relating the component to the party that made it. In order
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
A simple lattice-gas model for the electrocatalytic carbon monoxide oxidation on a platinum electrode is studied by dynamic Monte Carlo simulations. The CO oxidation takes place through a Langmuir–Hinshelwood reaction between adsorbed CO... more
A simple lattice-gas model for the electrocatalytic carbon monoxide oxidation on a platinum electrode is studied by dynamic Monte Carlo simulations. The CO oxidation takes place through a Langmuir–Hinshelwood reaction between adsorbed CO and an adsorbed OH radical resulting from the dissociative adsorption of water. The model enables the investigation of the role of CO surface mobility on the macroscopic electrochemical response such as linear sweep voltammetry and potential step chronoamperometry. Our results show that the mean-field approximation, the traditional but often tacitly made assumption in electrochemistry, breaks down severely in the limit of vanishing CO surface mobility. Comparison of the simulated and experimental voltammetry suggests that on platinum CO oxidation is the intrinsically fastest reaction on the surface and that CO has a high surface mobility. However, under the same conditions, the model predicts some interesting deviations from the potential step curre...
Research Interests:
Temperature programmed desorption of CO coadsorbed with atomic N on Rh(100), reveals both long- and short-range interactions between adsorbed CO and N. For CO desorption from Rh(100) at low coverage we find an activation energy Ea of... more
Temperature programmed desorption of CO coadsorbed with atomic N on Rh(100), reveals both long- and short-range interactions between adsorbed CO and N. For CO desorption from Rh(100) at low coverage we find an activation energy Ea of 137±2 kJ/mol and a preexponential factor of 1013.8±0.2 s−1. Coadsorption with N partially blocks CO adsorption and destabilizes CO by lowering Ea for CO desorption. Destabilization at low N coverage is explained by long-range electronic modification of the Rh(100) surface. At high N and CO coverage, we find evidence for a short-range repulsive lateral interaction between COads and Nads in neighboring positions. We derive a pairwise repulsive interaction ωCO–NNN=19 kJ/mol for CO coadsorbed to a c(2×2) arrangement of N atoms. This has important implications for the lateral distribution of coadsorbed CO and N at different adsorbate coverages. Regarding the different lateral interactions and mobility of adsorbates, we propose a structural model which satisf...
Research Interests:
We present a dynamic Monte-Carlo model involving lateral interactions and different adsorption sites (top, fcc and hcp). Using this model in combination with kinetic parameters from UHV experiments and lateral interactions derived from... more
We present a dynamic Monte-Carlo model involving lateral interactions and different adsorption sites (top, fcc and hcp). Using this model in combination with kinetic parameters from UHV experiments and lateral interactions derived from DFT calculations we have reproduced the ordering behavior of NO on Rh(111) during adsorption and the temperature programmed desorption (TPD) of NO from Rh(111) under UHV conditions. The formation of c(4×2)-2NO domains at 0.50 ML coverage is shown to depend strongly on the next-next-nearest-neighbor repulsion between the NO adsorbates in our model. The formation of the (2×2)-3NO structure at higher coverage follows from the avoidance of the strong next-nearest-neighbor repulsion in favor of the occupation of the top sites. A single-site model was able to reproduce the experimental TPD, but the lateral interactions were at odds with the values of the DFT calculations. A three-site model resolved this problem. It was found that all NO dissociates during ...
Research Interests:
ABSTRACT
Research Interests:
Abstract Smart spaces are physical spaces where services provided by Consumer Electronics (CE) devices with varying resource availabilities work together to realize user-specific automated scenarios. These scenarios may be interrupted in... more
Abstract Smart spaces are physical spaces where services provided by Consumer Electronics (CE) devices with varying resource availabilities work together to realize user-specific automated scenarios. These scenarios may be interrupted in case one of the services making up the scenario stops working, eg due to lack of resources, node failure or leave. Therefore, the user experience is highly dependent on the availability and reliability of smart space resources and services. This paper proposes an architecture for resource ...
Research Interests: Engineering, Computer Science, Architecture, User eXperience, Smart spaces, and 9 moreService management, Consumer Electronics, Service discovery, Use Case, Efficient Resource Management in Cloud Computing, Service Provider, service Business, Electrical And Electronic Engineering, and Resource Availability
The execution of a concurrent computation by a network of processors requires a routing algorithm that is deadlock free. Many routing algorithms proposed for processor networks have the potential of deadlock due to the cyclic topology of... more
The execution of a concurrent computation by a network of processors requires a routing algorithm that is deadlock free. Many routing algorithms proposed for processor networks have the potential of deadlock due to the cyclic topology of the network. In this paper we first formalize the concept of message routing. Next, we show a method by which a deadlock-free routing