Technologies

Digital Product Passports (DPPs) must preserve reliable component histories across repair, removal, and reuse, even when lifecycle updates are concurrent or reported retrospectively. This paper proposes an integrated model combining permanent identifiers, temporal containment relationships, causal versioning, and explicit acceptance decisions to ensure consistent lifecycle records. The model was implemented using Hyperledger Fabric, Java chaincode, and a Spring Boot REST API. Evaluation included blockchain-only and end-to-end tests covering concurrent lifecycle operations, transaction performance, historical reconstruction, and idempotency checks. The results confirmed correct lifecycle transitions and reconstruction of accepted product states. At 200 concurrent clients, sustained UPDATE and direct GET throughput reached 348.61 and 1210.98 requests/s, respectively. Across three contention experiments, 265 of 300 INSTALL requests were accepted, with no inconsistent component placements detected. On an expanded ledger, historical reconstruction averaged 1562.3 ms with one client and 4135.0 ms with ten clients, revealing a scalability limitation. All experiments were conducted on a single-machine Hyperledger Fabric prototype. Improved historical indexing and evaluation in physically distributed environments remain for future work.

In this study, the dynamic behaviour of a .223 Remington rifle barrel was investigated during live firing. A three-dimensional finite element modal analysis was performed and combined with experimental displacement measurements. The numerical model was used to determine and physically classify the first ten eigenmodes of the barrel–muzzle-device assembly according to their mode shapes and dominant deformation directions. The analyses revealed characteristic modal frequency regions at approximately 44.7 Hz, 259.8 Hz, 705–706 Hz, 1349–1356 Hz, and 2153 Hz. During the firing tests, experimental displacement time histories were measured at two locations along the barrel. The Fast Fourier Transform (FFT) was used to convert the time-domain signals into frequency-domain spectra. Experimental spectral peaks were identified independently of the numerical predictions using a local spectral-contrast criterion and subsequently compared with the modal frequency regions predicted by the finite element method (FEM). The most consistently identified experimental frequency components occurred near 270 Hz in the Front displacement spectra, with relative deviations of 3.72–5.30% from the FEM-predicted bending frequency of 259.81 Hz. In the higher-frequency regions, the correspondence between the experimental peaks and the numerical predictions was less consistent across the firing configurations and measurement points. The results demonstrate partial frequency correspondence between the experimentally measured barrel response and the numerical modal predictions. However, the presence of spectral peaks at comparable frequencies alone does not confirm the experimental identification of the corresponding mode shapes. The findings therefore highlight the importance of distinguishing between frequency correspondence and the independent experimental identification of individual vibration modes.

Driver drowsiness is a major cause of fatal car accidents, and camera-based monitoring offers a promising way to detect it. Detecting drowsiness from video requires modeling both facial appearance and its temporal evolution. In this article, we propose MSCT-Net, a multi-stream CNN–Transformer network for this purpose. Region-specific CNN backbones encode the face, right eye and mouth in every frame. Then, a single Transformer encoder jointly models their temporal dynamics across regions and time. MSCT-Net is trained from scratch on the challenging NTHU Drowsy Driver Detection dataset, which covers daytime and nighttime driving with and without eyewear. The architectures for the backbones and the Transformer are selected experimentally on the validation set. Compared with a baseline that shares the same backbones but averages features over time, MSCT-Net improves the clip-level accuracy from 78.70% to 82.16%, the F1-score from 81.15% to 83.15% and the AUC from 0.8807 to 0.9052, with higher accuracy and AUC in every scenario of the dataset. It also outperforms a CNN-LSTM baseline of similar size (accuracy 80.94%, F1-score 82.79%, AUC 0.8983). On the official test set of 14 subjects, MSCT-Net achieves a frame-level accuracy of 82.26% without any post-processing.