Articles & Research Papers

This study benchmarks three quantized convolution pipelines mapped to a mid-range Artix FPGA for on-device particulate matter classification. The student-led team optimized memory movement with line-buffer reuse and demonstrated a 38% reduction in dynamic power compared with baseline HLS-generated kernels while retaining 94.1% classification accuracy. Field tests in school-zone corridors indicate stable latency under variable thermal conditions, supporting practical deployment in battery-constrained sensing stations.

The paper introduces a staircase-herringbone channel geometry designed through CFD iterations and fabricated with two-stage soft lithography. Across six donor-matched blood samples, the geometry improved recovery of model tumor cells by 17% over a straight-channel control while maintaining viability above 92%. The authors discuss practical limits around throughput scaling and propose a modular cartridge framework for classroom-compatible translational prototyping.

This article develops a weighted graph planner that jointly minimizes travel time, acoustic exposure near schools, and emergency landing scarcity. Using open municipal GIS layers and measured propeller signatures, the model reduced aggregate neighborhood disturbance by 24% relative to shortest-path baselines in a Boston test district. The student researcher also reports a transparent policy dashboard that helps civic reviewers inspect route trade-offs before approval.

The authors evaluate a recycled PET-glass composite under one million load cycles to estimate service-life suitability for secondary bridge deck applications. Coupon and beam testing showed progressive stiffness degradation but no catastrophic fracture under 0.6 design stress ratio. A simple Miner-rule calibration, generated from undergraduate laboratory data, predicts inspection intervals within 8% of observed failure-onset trends.

This investigation compares two low-cost heat-recovery loops integrated into a bench-scale solar still suitable for pre-college laboratory settings. The best-performing loop increased freshwater yield from 2.8 to 3.6 L/m²/day during matched irradiance windows and reduced startup transients by recapturing condenser waste heat. The paper includes uncertainty analysis for salinity sensors and practical fabrication notes for reproducibility in school workshops.

The paper reports a lightweight occupancy estimator that leverages event-camera motion signatures rather than reconstructing identifiable frames. Evaluated across three campus corridors over four weeks, the method achieved a mean absolute error of 1.4 occupants and reduced personally identifiable data exposure compared with RGB baselines. The authors include deployment guidance for institutions balancing energy optimization with responsible sensing practices.

This article describes a low-cost analytics stack for tracking lithium-ion degradation in educational electric vehicle projects. The system combines impedance snapshots, temperature-normalized coulomb counting, and anomaly flags from a compact edge controller. Over a semester-long pilot, the model identified early cell imbalance events up to nine charge cycles before threshold alarms triggered in the stock battery management unit.

The study evaluates adaptive vibration cues for improving control consistency in myoelectric prosthetic training. A four-week protocol with twelve participants reduced target acquisition error by 19% compared with fixed-pattern feedback, with strongest gains in novice users. The paper contributes an open calibration routine and reflects on ethical considerations when collecting biometric interaction data from student participants.