Procomcure Biotech (Austria)

Abstract An approach is presented to the facile preparation of biointerface‐caped periodic arrays of gold nanoparticles (Au NP) with controlled plasmonic and advanced sensing properties. It relies on feedback‐controlled UV‐laser interference lithography combined with a novel ion milling method. These advancements improved preparation reproducibility yielding localized surface plasmon resonance (LSPR) wavelength tuned to near infrared spectrum with standard deviation of 10 nm. Moreover, a novel molecular toolkit of thiols and silatrane linkers with carboxybetaine/sulfobetaine headgroups is employed for orthogonal chemical modification of Au NPs and passivation of the optically non‐active sensor chip regions, to prevent non‐specific sorption of molecules from liquid samples. As part of this toolbox, the use of mercaptopropyl silatrane is reported as an adhesion‐promoting coating for attaching Au NPs to an oxide substrate, replacing the traditionally used strongly absorbing chromium. The developed biofunctional optical nanostructures show LSPR with improved quality factor and enable 2.5‐fold increase in accuracy for tracking of biomolecular binding on their surface via LSPR wavelength variations monitoring. In addition, a strategy is demonstrated to mitigate unspecific biomolecule sorption not only at the plasmonic hotspot on the metal surface, but also on the often‐overlooked oxide substrate, highlighting the importance of the described orthogonal modification.

Modern embedded automotive software uses AUTOSAR for software development. This software is organized into a set of runnables that represent basic functionality. For optimal resource utilization, runnables are grouped into tasks. In AUTOSAR, a schedule table is used for the deterministic time triggering of task activations or events. In the state of the art, the schedule table is generated at design time. The memory demand of the schedule table is highly sensitive to the application parameters, i.e., the periods and offsets of the runnables and task types. Variations in these parameter values can very significantly increase the memory demand of the schedule table, even for small applications consisting of, e.g., only four runnables. In this paper, we propose an alternative approach for the online generation of the schedule table with an upper bound on the memory demand that is insensitive to variations in the values of the periods and offsets of runnables, and task types. Using multiple case studies, we show that our approach, despite a slight runtime overhead, significantly reduces memory demand.

ABSTRACT This paper addresses the challenge of automating process‐centric quality assurance (QA) in safety‐critical domains, where compliance with regulations is crucial. Currently, QA engineers manually check compliance using tedious methods like browsing engineering artifacts and ad‐hoc scripts. Automated support could improve efficiency, but it requires constraints to be written in structured, executable forms (e.g., in the Object Constraint Language, OCL), whereas engineers prefer natural language. To bridge this gap, we propose the use of large language models (LLMs) to generate OCL from natural language, enhanced by schema‐based prompting and domain‐specific language (DSL)–based repairs. Unlike prior work focused on UML models, this work applies OCL to software process QA. Evaluating six LLMs, we find o1‐mini and Codestral perform best, with our automatic repairs ensuring constraint executability for 22%–44% of an LLM's generated OCL constraints that would otherwise remain nonexecutable due to errors.

Traditional embedded Real-Time Operating Systems (RTOS) or Basic Software (BSW) implementations typically require manual porting to new hardware platforms. However, this approach can be time-consuming and error-prone, especially given the frequent introduction of new or upgraded hardware architectures. In addition, traditional testing methods may not fully capture the complexity and nuances of the system, making it difficult to ensure correctness and dependability. To address these challenges, we propose a comprehensive methodology that integrates formal methods, a WCET tool, and a code generation technique. We use formal methods to create models and verify their correctness against functional and non-functional specifications or properties such as safety, liveness, and timing. We use the WCET tool as a microarchitecture analyzer to analyze the low-level binary code. The intermediate results of the tool are used to verify the correctness of the software implementation against the runtime effects of the hardware, such as data/memory race conditions. Finally, the code is generated from the formal models. Our proposed framework simplifies the maintenance of RTOS or BSW implementations while ensuring their correctness and partially automating portability to new hardware architectures.

Learning underlying patterns from sensory data is crucial in the Human Activity Recognition (HAR) task to avoid poor generalization when coping with unseen data. A key solution to such an issue is representation learning, which becomes essential when input signals contain activities with similar patterns or when patterns generated by different subjects for the same activity vary. To address these issues, we seek a solution to increase generalization by learning the underlying factors of each sensor signal. We develop a novel multi-channel asymmetric auto-encoder to recreate input signals precisely and extract indicative unsupervised futures. Further, we investigate the role of various activation functions in signal reconstruction to ensure the model preserves the patterns of each activity in the output. Our main contribution is that we propose a multi-task learning model to enhance representation learning through shared layers between signal reconstruction and the HAR task to improve the robustness of the model in coping with users not included in the training phase. The proposed model learns shared features between different tasks that are indeed the underlying factors of each input signal. We validate our multi-task learning model using several publicly available HAR datasets, UCI-HAR, MHealth, PAMAP2, and USC-HAD, and an in-house alpine skiing dataset collected in the wild, where our model achieved 99%, 99%, 95%, 88%, and 92% accuracy. Our proposed method shows consistent performance and good generalization on all the datasets compared to the state of the art.

Many long-established, traditional manufacturing businesses are becoming more digital and data-driven to improve their production. These companies are embracing visual analytics in these transitions through their adoption of commercial dashboarding systems. Although a number of studies have looked at the technical challenges of adopting these systems, very few have focused on the socio-technical issues that arise. In this paper, we report on the results of an interview study with 17 participants working in a range of roles at a long-established, traditional manufacturing company as they adopted Microsoft Power BI. The results highlight a number of socio-technical challenges the employees faced, including difficulties in training, using and creating dashboards, and transitioning to a modern digital company. Based on these results, we propose a number of opportunities for both companies and visualization researchers to improve these difficult transitions, as well as opportunities for rethinking how we design dashboarding systems for real-world use.

Interdiffusion – defined as the mutual interpenetration of macromolecules to form a diffuse interphase – is observed at polymer-polymer interfaces in a wide range of plastics processing techniques. This transport phenomenon has been studied extensively on static interfaces (i.e., the polymers and, in particular, the interface are not subject to deformation), for instance, in plastics welding, and consolidation of thermoplastic UD tapes. In contrast, interdiffusion under pressure-flow conditions – as present in stratified flows during co-extrusion – has been studied scarcely to date. In particular, the correlation between the magnitude of interfacial shear load and the rate of interdiffusion is unknown. In this work, a two-layer co-extrusion demonstration die and the material combination poly(methyl methacrylate) (PMMA) and styrene-co-acrylonitrile copolymer (SAN) were used to study the effect of the interfacial shear stress on the rate of interdiffusion. The implementation of a model-based digital process twin enabled to accurately adjust the co-extrusion processing conditions. Employing a partial factorial experimental design, the magnitude of interfacial shear stress was varied at certain levels of interfacial contact time and melt temperature. The diffuse interphase of the calibrated two-layer sheets was then characterized by means of confocal Raman microscopy and an apparent mutual diffusion coefficient was determined. It is shown that already low levels of interfacial shear stresses (e.g., 1,000 Pa) significantly boost the rate of interdiffusion compared to static conditions. Though, a gradual increase of the interfacial shear stress level up to approximately 40,000 Pa revealed no further impact on the mutual interdiffusion coefficient.

Guided Tour for Dashboard Onboarding

Drone delivery has been one of the promising applications of drones in real life. Precision Landing is an important mission to deliver the package to the customer’s place. Visual marker has been studied to provide accuracy in pose estimation during landing based on a downward camera since GPS localization may be disrupted at low altitude due to high building occlusions. However, there are still some remaining issues of previous visual marker design such as the trade-off between detection altitude and marker size, lack of redundancies, weakness to motion blur, occlusions on the landing pad and possible issues related to the utilization of a single marker family. This paper proposed a multiple marker design for precision and redundant visual landing combining multiple types of marker families: Whycon, Aruco and AprilTag. The design and respective algorithms to detect multiple markers on the design have been evaluated in both photo-realistic simulation and practical environments. A smooth and precise trajectory for visual landing is successfully performed from the altitude of 25 m even in the case of occlusions or possible issues happened with one of the marker detector algorithms.

This paper presents a novel approach for counting hand-performed activities using deep learning and inertial measurement units (IMUs). The particular challenge in this task is finding the correct window size for capturing activities with different durations. Traditionally, fixed window sizes have been used, which occasionally result in incorrectly represented activities. To address this limitation, we propose segmenting the time series data into variable-length sequences using ragged tensors to store and process the data. Additionally, our approach utilizes weakly labeled data to simplify the annotation process and reduce the time to prepare annotated data for machine learning algorithms. Thus, the model receives only partial information about the performed activity. Therefore, we propose an LSTM-based architecture, which takes into account both the ragged tensors and the weak labels. To the best of our knowledge, no prior studies attempted counting utilizing variable-size IMU acceleration data with relatively low computational requirements using the number of completed repetitions of hand-performed activities as a label. Hence, we present the data segmentation method we employed and the model architecture that we implemented to show the effectiveness of our approach. Our results are evaluated using the Skoda public dataset for Human activity recognition (HAR) and demonstrate a repetition error of ±1 even in the most challenging cases. The findings of this study have applications and can be beneficial for various fields, including healthcare, sports and fitness, human-computer interaction, robotics, and the manufacturing industry.