Publications

GT-BoW is a geometry-aware image-retrieval method for ground-texture localization. It combines an approximate k-means vocabulary and soft assignment with constraints derived from the fixed scale and consistent orientation of downward-facing images. Separate high-accuracy and high-speed variants address the different demands of global localization and loop-closure detection. On the full benchmark, the method increases global-localization mean average precision from 0.026 to 0.559 and detects nearly 3× as many loop closures as a standard DBoW baseline.

Pruning weight helps growers assess vine vigor and plan vineyard management, but conventional measurement methods are labor-intensive or require specialized equipment. This work presents an affordable computer-vision method that combines a smartphone camera with structured light. It achieves R² = 0.80 on vertical-shoot-positioned vines, outperforming prior vision-based approaches, and reports initial results (R² = 0.29) for high-cordon vines, a training system not previously evaluated with computer vision.

L-GROUT is a lightweight global-localization system that matches natural ground texture observed by a downward-facing camera to a prebuilt map. Its main contributions are improved database-feature selection, locality-preserving dimensionality reduction for fast binary features, and a spatial-voting stage that retains accuracy in lightweight configurations. The complete system produces millimeter-scale estimates at more than 4 Hz on a Raspberry Pi 4, without GPU acceleration or environmental modification.

Cold, remote, and extraterrestrial environments make conventional robot construction and repair difficult, but they may provide abundant water ice. This work demonstrates modular robots, manipulators, rovers, and soft robots that use ice as a structural material or exploit freezing, melting, and fracture to change their capabilities. A scaling analysis examines how ice structures perform across different sizes and ambient temperatures.

This book chapter reviews wearable chemical sensors for health monitoring, including approaches for sampling and analyzing biofluids and the design considerations involved in translating these sensors into continuous, non-invasive wearable systems.

This review surveys emerging wearable technologies for personalized health and performance monitoring, with an emphasis on sensing platforms that can collect physiological and biochemical measurements continuously outside conventional clinical settings.