This paper investigates an alternative depth-sensing method based on the distance-dependent irradiance fall-off of two geometrically distinct light sources, i.e. a point source and a line source. A corresponding distance model is derived and a first experimental proof of concept using low-cost hardware is presented. The model expresses object distance as a function of the ratio between measured sensor signals under both illumination conditions. In contrast to purely intensity-based methods, a key advantage of the proposed concept is that the distance estimate is independent of the absolute reflectance of a Lambertian target. To assess practical feasibility, an experimental system is implemented using off-the-shelf components. After calibration, the system is capable of estimating the distance of objects with different reflectances. The results indicate that the method can provide useful distance estimates within a limited operating range.
Radio Frequency Identification (RFID)-based gesture recognition is affected by domain shifts caused by changes in sensing distance and environment. We propose a supervised cross-domain adaptation framework that combines a graph convolutional network (GCN) backbone with domain-adversarial neural network (DANN) training for gesture recognition using body-worn RFID tags. During adaptation, labeled gesture samples from the source and both target training sets are used, while a Gradient Reversal Layer encourages domain-invariant representations. Over five training seeds, the model achieves accuracies of 91.58±0.39%, 94.43±0.83%, and 87.34±1.42% in three environments with various distance between the antenna and the person performing the gestures. The proposed framework has practical applications in gesture-based human-computer interaction, hands-free control in smart homes and industrial settings, and movement monitoring for healthcare and rehabilitation
This paper presents a wireless sensing system to monitor the structural integrity of magnesium-silver alloy based biodegradable sutures, specifically designed for sternal applications. A body-matched ultra-wideband coplanar waveguide-fed circular disc monopole antenna is designed to propagate into the complex lossy medium of the human torso. It detects structural changes in a loop-shaped suture by monitoring variations in return loss characteristics, enabled by coupling between the antenna and the suture. The system performance is validated via simulations and measurements with numerical and physical phantoms, including analyses for suture-antenna misalignment and suture depth. Moreover, the on-body antenna successfully monitors accelerated suture degradation in an acidic tissue-mimicking phantom, simulating long-term effects. It has been shown that degradation can be detected up to a depth of 20 mm in the body. This approach enables non-invasive, real-time feedback on sternal healing, potentially improving post-operative care and patient outcomes.
In this work, a smartphone-based, compact, and time-resolved luminescence (TRL) microscopic imager with 3D-printed assembly and a timing-tunable synchronization circuit was developed. Using water soluble Mn-doped quantum dots (QDs) as long-lived luminescent probes, the imager was optimized to achieve a high signal/background ratio through tuning the timing between the optical chopper and LED excitation. Under optimized time-resolved conditions, the imager achieved a limit of detection (LOD) of 0.05 μg/mL on detecting the probes, which is around 70-fold lower when compared with a LOD of 3.5 μg/mL under non-resolved imaging. The imager was further applied to Cu²⁺ detection in water, achieving a LOD of 0.05 μM and a selective response against the tested interfering metal ions commonly found in water samples. The imager should have broader applications in biosensing/imaging.
A high-precision platinum Resistance Temperature Detector (RTD) based on a Positive Temperature Coefficient (PTC) sensing mechanism was developed for home appliance applications, using statistical simulation and a dual-stage adjustment strategy to significantly improve fabrication yield and resistance control. The proposed approach addresses process-induced variability by employing a dual resistance-tuning technique that enables precise fine-tuning within a single device architecture. Laser trimming parameters were systematically optimized to ensure clean, controlled ablation of platinum-coated sapphire while minimizing thermal and mechanical damage. The design incorporates twenty-one quantized trimming links (seven each of 15, 10, and 5 Ω) for resistance increase, alongside a continuously scalable subtractive bypass defined via maskless direct-write lithography for resistance reduction. This hybrid adjustment scheme compensates for both under- and over-target resistance variations, resulting in a tightly controlled 1000 Ω target across the wafer (standard deviation reduced from 23.6 Ω to 1.8 Ω), with enhanced yield and improved device uniformity. Fabrication was carried out using mask aligner-based lithography, followed by deposition of a 15 nm Titanium (Ti) adhesion layer and a 300 nm platinum (Pt) conductive layer. Rapid Thermal Processing at 700 °C for 60 s under a nitrogen atmosphere further improved thin-film crystallinity and electrical stability. The combined design and process strategy provides a scalable, robust method for producing uniform, stable, high-yield RTD sensors suitable for large-scale manufacturing.
