Eglin Air Force Base

Active and efficient control of nonlinear optical processes is essential for integrated photonics, with applications in signal processing, ultrafast switching, and quantum light manipulation. While nanophotonic structures are powerful for enhancing nonlinearities, achieving wide-range electrical tunability has remained a challenge. Plasmonic tunnel junctions offer a unique path to bridge this gap because they combine extreme optical field confinement with direct electrical control in a single nanoscale device. Here, we report the first, to our knowledge, demonstration of electrically tunable second-harmonic generation (SHG) in plasmonic tunnel junctions. Using ultra-stable epitaxial heterostructures, we achieve reproducible modulation of SHG with depths up to ∼500% and magnitudes above 1.3V −1 . We identify two mechanisms, electric-field-induced SHG (EFISH) and ion migration, that can either compete or cooperate depending on junction thickness and bias, enabling both broad tunability and ferroelectric-like hysteretic switching. These findings establish plasmonic tunnel junctions as a platform for electrically controlled nonlinear optics, with potential for nanoscale light sources, reconfigurable modulators, and neuromorphic photonic devices.

The ultrafast and large-contrast switching of light is essential to the realization of high-speed, energy-efficient optical logic elements and sources. Such functionality requires optical materials or systems whose optical response can be rapidly and dramatically tuned under an external stimulus. However, achieving high-contrast switching, in sub-picosecond time scales, with low power over macroscopic length scales, is exceedingly difficult. We utilize self-assembled, cavity-coupled tin-doped indium oxide nanocrystal monolayers to achieve sub-picosecond, high-contrast switching across the mid-infrared (from 2.5 to 6.1 μm). We achieve reflection modulation up to 40% absolute and 1200% relative depth, under moderate fluences of a few hundred μJ/cm2, whereas our pump-probe experiments reveal an upper bound on response times of 210 femtoseconds. The strong and ultrafast nonlinearity, along with the broad spectral tunability and scalable fabrication method, make the demonstrated system an appealing platform for a range of mid-infrared applications, including ultrafast optical switching and mode-locked lasers.

Autonomous UAVs operating in GPS-denied environments rely heavily on onboard perception for safe navigation, yet traditional LiDAR and camera-based systems falter in poor lighting and weather situations. In contrast, millimetre wave (mmWave) radar offers superior robustness and low power consumption but produces extremely sparse, noisy, and low-resolution measurements that limit its utility for dense 3D understanding. This work proposes a novel radar-only inference framework that reconstructs dense LiDAR-like point clouds by leveraging a stereo millimeter-wave radar (IWR6843AoP) data with hallucinated Time-of-Flight (ToF) representations. Rather than relying on explicit ToF sensors, our method learns to infer mid-range depth priors through an auxiliary supervised hallucination network trained with radar inputs only. A unified PointNet-based architecture fuses stereo radar features with the hallucinated ToF embedding to generate dense Cartesian point clouds in real time. A large-scale, high-fidelity multimodal dataset of approximately 50,000 frames is generated in NVIDIA Isaac Sim to emulate realistic UAV sensor configurations. Experimental evaluations demonstrate strong robustness to radar noise and reliable geometric consistency suitable for downstream tasks such as ICP-based odometry. The resulting framework delivers LiDAR-like geometric structure using only radar at runtime, offering a computationally efficient, deployable, and resilient solution for UAV navigation in GPS-denied and harsh environmental conditions.

Low-cost, low-power sensors are essential for embedded and edge applications such as miniature Unmanned Aerial Vehicles (UAV) and robotics, but their sparse, noisy, and low-fidelity outputs limit their effectiveness in perception tasks. We propose an AI-driven framework that enhances these sensors by transforming low-resolution measurements into dense surface representations. Our approach learns the unsigned distance field anchored by noisy depth measurements to extract predicted truth surface. To enable robust testing, we use simulation and digital twins to generate synthetic data across diverse environments. Experiments show that dense reconstructions from our approach can robustly improve downstream tasks such as navigation, and extends the utility and resilience of low-power sensor systems in resource-constrained environments.

There are no established family medicine (FM) residency-level global health objectives or curricula for program directors who want to offer ethical and meaningful global health experiences to residents. We sought to develop and evaluate a toolkit of resources that family medicine educators can use to improve residents' knowledge of this important aspect of family practice. We reviewed and categorized peer-reviewed and grey literature publications and global health curricula from FM and non-FM residency programs related to clinical topics, standards of global health practice and partnerships, decolonization, and others. The toolkit uses six standard Accreditation Council for Graduate Medical Education (ACGME) competencies to organize learning objectives, content areas, and resources. Resources were further categorized based on complexity, cost, and time. We developed the toolkit with a focus on patient safety and ethical engagement with global partners. We evaluated the toolkit using an online survey of global health educators from inside and outside the United States. The toolkit was vetted by the Society of Teachers of Family Medicine Board of Directors and published as an open-access resource on the STFM website (https://stfm.org/teachingresources/curriculum/globalhealthtoolkit/overview/).The toolkit had 1,446 unique views in the 18 months after it was published. Most global health educators surveyed found the toolkit to be appropriate for resident-level education and ethically sound. Respondents indicated that they would most likely use the toolkit to improve an existing global health experience or track. A novel toolkit provides resources curated and presented by topic, complexity, and estimated cost that residency programs can use to create or augment global health offerings. Educators can use resources and associated learning objectives presented using standard ACGME competencies to connect knowledge, skills, and attitudes gained through global health education to their residents' overall learning needs.

A nonlinear 3-D navigation algorithm based on a generalization of the classical law by Adler is developed. In the development of the engagement kinematics it is shown that choosing two opposite directions of the unit normal to the lead angle plane formed by the missile velocity and the line of sight (LOS), and two different roots of a quadratic equation associated with the true collision velocity, one arrives at four different scenarios which give rise to distinct missile trajectories and distinct interception times. Through numerical simulation, it is shown that one of the choices leads to a missile trajectory that deviates from the target, while the other three result in interception of the target. The generalization of Adler's proportional navigation comprises two equations, a linear one and a nonlinear relation that possesses the turning rates of the LOS, the navigation gain, the lead angle, the range, and a velocity correction that brings the instantaneous collision-course velocity vector on the locus of the missile velocity. Numerical tests based on the generalized 3-D navigation scheme and Adler's proportional navigation law are compared and discussed. The generalized law is found to require less acceleration for intercept.

The target tracking problem is addressed for multi-agent systems where the target state information is only partially available to the agents via a heterogeneous measurement model. A necessary and sufficient condition, termed trackability, is provided, to indicate the feasibility for tracking a target with partial measurements. A Lyapunov-based deep neural network (Lb-DNN) adaptive controller is developed to achieve target tracking, under the trackability condition, by adaptively compensating for the uncertainty stemming from the unknown target dynamics. A Lyapunov-based stability analysis is provided to guarantee exponential target state estimation and tracking within a neighborhood of the target state.