Industries

Project CIPACTLI (Compact Integrated Polymer Amphibian for Coastal and Littoral Investigation) evaluated whether rare-earth neodymium (N52) magnetic couplings provided an operational advantage over flexible magnetic tape in a rapid-deployment, cannon-launched amphibious robotic chassis intended for search-and-rescue (SAR) and environmental-monitoring use. Two prototype iterations were tested across a towed transit-rate proxy for movement performance, static structural loading, and ballistic-launch survivability. The first prototype was a tensegrity-inspired, segmented soft robot with inflatable buoyancy modules joined by magnetic couplings; a 2³ factorial design varied core tendon material (nylon vs. steel), substrate (land vs. water), and magnet type (flexible tape vs. N52 neodymium). Tendon material was the dominant driver of transit rate (partial η² = 0.637, p < 0.001); magnet type showed no significant main effect (p = 0.507). Because the per-cell sample size (n = 3) yielded low statistical power (~0.15), the null result was additionally evaluated with a two one-sided-tests equivalence procedure at a pre-specified margin of ±0.20 m/s (≈18% of the grand mean); the reconstructed 90% confidence interval for the magnet-type difference (−0.067 to +0.152 m/s) fell entirely inside this margin, supporting statistical equivalence at that margin, though a narrower margin could not be resolved with the present sample. Both magnet types achieved 3/3 survival under ballistic launch. The second prototype eliminated magnetic couplings and inflatable bladders in favor of a monolithic PLA body with integrated buoyancy; this design change (which altered four construction variables simultaneously and therefore cannot isolate any single feature’s contribution) was associated with a 6.96-fold increase in quasi-static compressive load capacity (p = 0.001), equivalent ballistic survivability, and a 24.9% reduction in towed aquatic transit rate. Tethered-deployment trials (n = 6 per condition) showed no observed tether-related failures. Within this bounded operational envelope—one magnet geometry and grade, one mounting scheme, one energy level, ambient chlorinated fresh water, and room temperature—rare-earth N52 magnets conferred no measurable transit rate or survivability advantage over flexible magnetic tape. Corrosion in saline water, thermal cycling, and long-duration exposure, which govern neodymium performance in actual maritime service, were not evaluated and are identified as priority future work before any broader material-substitution recommendation is warranted.

Medical ultrasound transducers require acoustic impedance matching layers to bridge the gap between piezoelectric ceramics (~30–38 MRayl) and biological tissue (~1.5–1.7 MRayl). Existing solutions—such as alumina–epoxy, anodic aluminum oxide–epoxy 1–3 composites, magnesium alloys, and gradient nanocomposites—rely on high-temperature curing, precision machining, or multi-step fabrication, which increase production costs and limit design flexibility. Viscoelastic suspensions offer a potential low-cost alternative, but prior to any clinical assessment, manufacturing protocols must be bounded at the exploratory stage. This study does not demonstrate impedance matching, quarter-wavelength performance, or immediate clinical relevance. This work defines a Stage 1 pre-pilot (Technology Readiness Level 2–3) exploratory study to bound manufacturing protocols, rank formulations, and identify candidate parameters for future Quality by Design implementation. Six formulations were tested in open-cell polyurethane foam matrix: thick cornstarch–water (50:50 wt%), thin cornstarch–water (66:33 wt%), polystyrene sulfonate (50:50 wt%), polyethylene glycol–fumed silica (40:60 wt%), boric acid–simethicone, and a boric acid–cornstarch hybrid. The 5 mm prototypes function as bulk composite samples rather than functional matching layers, and all testing was performed at audio frequencies (50 Hz–20 kHz) with uncalibrated consumer-grade instrumentation. At 50–1000 Hz, boric acid–simethicone exhibited the lowest attenuation (32 dB). At 10,000–20,000 Hz, polystyrene sulfonate and thin cornstarch showed the lowest attenuation (97 dB at 20 kHz). Attenuation increased with frequency; a dip between 100 and 1000 Hz was attributed to fixture interference. Two independent recordings were obtained per condition (n = 2). No inferential statistics, confidence intervals, or variance estimates were computed. The results establish a descriptive baseline but do not demonstrate impedance matching or immediate clinical relevance; the 5 mm prototypes function as bulk composite samples rather than quarter-wavelength matching layers. Should subsequent high-frequency validation prove promising, future work includes factorial or definitive screening designs, hydrophone calibration per international standards, Z = ρc measurements, substrate deconvolution, 0.5–1.0 mm layers for 650 kHz, phantom validation, and ISO 10993 biocompatibility screening.

Automated visual inspection has become a critical component of Industry 4.0 manufacturing systems, enabling improved product quality, reduced operational costs, and increased production efficiency. In the carpet manufacturing industry, quality inspection remains particularly challenging due to the high variability of textures, colors, patterns, and defect appearances, which often limits the effectiveness of conventional inspection systems. Furthermore, the scarcity of publicly available industrial carpet defect datasets limits the realistic assessment of automated inspection methods. Existing approaches are therefore frequently evaluated on benchmark datasets that do not fully capture the variability of real manufacturing environments or rely on computationally demanding end-to-end deep learning architectures, which may reduce their suitability for practical industrial deployment. To address these challenges, this paper introduces a new industrial carpet defect dataset acquired under controlled imaging conditions and containing defect-free samples together with four representative defect categories: Backing Seam, Bad Mend, Dirty Face, and End Out. In addition, we present an engineering study that systematically evaluates a practical two-stage inspection framework combining pretrained deep feature extraction, one-class anomaly detection, and supervised defect classification for industrial carpet quality control. A distinguishing aspect of this study is the systematic evaluation of the framework on both the public MVTec AD Carpet benchmark and the newly introduced industrial dataset, enabling its performance to be assessed under both benchmark and realistic industrial conditions. The first stage performs anomaly detection using one-class learning models trained exclusively on defect-free samples, while the second stage classifies detected defects into predefined categories. To investigate the impact of visual representations on inspection performance, multiple deep feature extraction backbones, including ConvNeXt, VGG-19, and Wide ResNet-50-2, are evaluated in combination with lightweight machine learning models. Experimental results demonstrate strong inspection performance, achieving an AUROC of 99.92% on the MVTec AD Carpet benchmark and 93.92% on the proposed industrial carpet dataset. The obtained results demonstrate that this integration of established techniques provides an effective and computationally lightweight solution for industrial deployment, while confirming the practical value of the proposed dataset as a realistic benchmark for future research.