Distributed energy systems increasingly consist of heterogeneous assets and organizations that must exchange operational data while preserving interoperability, security, and regulatory compliance. Existing integration solutions often rely on syntactic adapters or centralized data hubs, which scale poorly and offer limited transparency or governance. This paper presents a metadata-driven federated monitoring architecture that integrates ontology-based metadata federation, event-driven microservices, and governance-aware provenance tracking to enable secure, scalable, and auditable data sharing across distributed energy infrastructures. The proposed system models all assets and data streams through a unified semantic graph, aligning heterogeneous schemas via automated ontology matching and combined lexical–structural similarity scoring. A microservices pipeline ingests multi-protocol data (OPC-UA, MQTT, REST), applies stream analytics for anomaly detection, and enforces access and compliance policies at the metadata layer. A Web-based interface allows operators to issue GraphQL queries, visualize distributed assets, and monitor real-time alerts linked to provenance records. A prototype implementation demonstrates operational-scale efficiency, achieving low-latency response (≤540 ms for hybrid metadata–telemetry queries over 10,000 assets), near-linear scalability (∼4.5% CPU growth per added node), and high governance accuracy (precision 0.90, recall 0.95, median detection 1.6 s) while maintaining minimal overhead (<8% added latency). These results highlight that the proposed metadata-driven federation delivers both technical performance and governance reliability unmatched by existing Web-based integration frameworks. These results show that metadata federation can be deployed at operational scale while providing explainable compliance and trustworthy data sharing across organizational boundaries. This research advances the state of the art in Web-based system engineering by combining semantic modeling, distributed processing, and security governance into a single deployable framework. Beyond energy systems, the approach offers a foundation for interoperable and auditable monitoring in other critical cyber-physical domains such as industrial IoT, urban infrastructure, and healthcare telemetry.

Ensuring data-flow integrity and rapid threat containment in renewable-integrated, distributed energy systems requires monitoring solutions that are technically rigorous yet lightweight in operation. This paper presents a service-oriented web framework for real-time data-flow tracing and threat propagation analysis in heterogeneous industrial control and energy networks. The framework integrates lightweight provenance tokens embedded in event streams, an incrementally maintained lineage graph with probability-weighted edges, and propagation-aware risk indicators that drive adaptive response orchestration through open web APIs. A progressive web dashboard provides sub-second visualization of dynamic topologies, risk heat maps, and operator controls. Implemented on a Kafka/Flink streaming backbone with a graph database and deployed in an eight-node Kubernetes testbed emulating substations, gateways, and adversarial nodes using OPC UA, MQTT, and REST, the system achieved tracing coverage of 0.96 ± 0.02 and fidelity of 0.92 ± 0.03, with forward propagation prediction reaching precision 0.91 and recall 0.88, outperforming static-topology baselines. Adaptive containment reduced the flow reproduction factor from 1.42 to 0.64, achieved a median containment efficacy of 0.71, and stabilized risk trajectories within two minutes, while operational cost remained low with payload expansion under 12%, CPU overhead below 4%, and service availability above 0.99 for critical assets. User studies showed 38% faster incident response and higher comprehension and confidence compared with static log viewers. These results demonstrate that modern web-engineering practices such as microservices, event-driven streaming, and progressive web interfaces can enable practical, real-time cyber defense for distributed energy infrastructures by bridging static security guidelines with deployable, adaptive situational awareness and containment.

Emerging automotive applications such as in-vehicle Augmented Reality (AR) and fully automated parking require a comprehensive understanding of the vehicle’s three-dimensional surrounding represented as anobject-levelenvironmental model. In this model, not only 3D poses (positions and orientations) and 3D sizes of detected objects are registered, but 3D shapes (geometries) need to be reconstructed precisely. A combination of 3D object detection and 3D surface reconstruction techniques, referred to asobject-level 3D reconstruction, is fundamental to building such environmental models. However, the possibilities to incorporate object-level 3D reconstruction in a car have not been sufficiently explored either in academic research or in the industry. This primarily stems from the cost and resource constraints associated with the automotive domain. In this paper, we address these constraints by proposing implementations ofin-vehicle object-level 3D reconstructionin two specific use cases:(i)augmented reality and(ii)automated parking. For augmented reality, we propose a cost-efficient solution called monocular3D Shapingthat requires only a single frame from a monocular camera as input. For automated parking, we propose a resource-efficient alternative that generates more precise 3D reconstruction results by taking advantage of additional 3D sensors (such as Lidars). The crux of our proposed approaches lies in the use of aLatent Shape Space, where various 3D shapes are represented using only two parameters. As a result, highly complex 3D shapes can now be transmitted using a low- to medium-bandwidth in-vehicle communication infrastructure in a cost-effective manner.

To provide safe autonomous driving to customers, the automotive industry faces a huge challenge to test and validate the self-driving functions. It is estimated that self-driving cars will cover about 240 million real or virtual kilometers on the journey to being mass production-ready. In other words, each function in the sense-plan-act loop of autonomous driving has to undergo a stringent test and validation process using Generationa large amount of real and simulated data. Thus, the generation of reference data for test and validation purposes at an industrial scale is a crucial topic and requires a high degree of automation. In this paper, we propose a deep learning-based approach named RoadNet to generate reference road data for testing and validating the in-vehicle lane detection function. Furthermore, we present our processing pipeline running in a data center which fully automatizes the generation of reference data through RoadNet. We evaluate the accuracy and generalization ability of the proposed method using the automation pipeline. Although improving the quantitative evaluation results is still work in progress, we believe that our solution to productionize a deep neural network is of great importance for the cost-sensitive automotive industry.

Future self-driving cars must be able to perceive and understand their surroundings. Deep learning based approaches promise to solve the perception problem but require a large amount of manually labeled training data. Active learning is a training procedure in which the model itself selects interesting samples for labeling based on their uncertainty, with substantially less data required for training. Recent research in active learning has mostly focused on the simple image classification task. In this paper, we propose novel methods to estimate sample uncertainties for 2D and 3D object detection using Ensembles. We moreover evaluate different training strategies including Continuous Training to alleviate increasing training times introduced by the active learning cycle. Finally, we investigate the effects of active learning on imbalanced datasets and possible interactions with class weighting. Experiment results show both increased time saving around 55% and data saving rates of around 30%. For the 3D object detection task, we show that our proposed uncertainty estimation method is valid, saving 35% of labeling efforts and thus is ready for application for automotive object detection use cases.

We propose a new approach called LiDAR-Flow to robustly estimate a dense scene flow by fusing a sparse LiDAR with stereo images. We take the advantage of the high accuracy of LiDAR to resolve the lack of information in some regions of stereo images due to textureless objects, shadows, ill-conditioned light environment and many more. Additionally, this fusion can overcome the difficulty of matching unstructured 3D points between LiDAR-only scans. Our LiDAR-Flow approach consists of three main steps; each of them exploits LiDAR measurements. First, we build strong seeds from LiDAR to enhance the robustness of matches between stereo images. The imagery part seeks the motion matches and increases the density of scene flow estimation. Then, a consistency check employs LiDAR seeds to remove the possible mismatches. Finally, LiDAR measurements constraint the edge-preserving interpolation method to fill the remaining gaps. In our evaluation we investigate the individual processing steps of our LiDAR-Flow approach and demonstrate the superior performance compared to image-only approach.

This paper presents a novel approach to reconstructing the 3D shape of an object from a single image. The approach combines deep neural networks with a silhouette-based 3D reconstruction process. The optimal 3D shape is sought efficiently inside an extremely low-dimensional latent shape space, and the viewpoint and the object shape are jointly optimized based on the result of image segmentation. Evaluation of this approach shows a nearly 20 percent performance gain in viewpoint estimation subsequent to the optimization.

We have experienced rapid development of augmented reality (AR) systems and platforms in the automotive industry. However, to bring AR into production cars, we still face a range of challenges to design an AR system that meets vehicle specific requirements. Based on our experience with an AR prototype car, we analyze the influence of augmented reality on the design of the in-vehicle electric/electronic (E/E) architecture.

In the last three years, a number of automotive Augmented Reality (AR) concepts and demonstrators have been presented, all looking for an interpretation of what AR in a car may look like. In October 2013, Mercedes-Benz exhibited to a public audience the AR In-Vehicle Infotainment (AR-IVI) system aimed at defining an overall in-vehicle electric/electronic (E/E) architecture for augmented reality rather than showing specific use cases. In this paper, we explain the requirements and design decisions that lead to the systemdesign, and we share the challenges and experiences in developing the AR-IVI system in the prototype vehicle. Based on our experiences, we give an outlook on future software and E/E architectural challenges of in-vehicle augmented reality.