The potential of uncrewed and autonomous vehicles in the air and on the ground is enormous. During my graduate studies I have gained expertise in developing autonomous UAS and UGV platforms. UASs and UGVs are used in the field of various research fields ranging from agriculture to defense. The autonomous feature of the uncrewed systems is the most important feature that has been utilized heavily. Another important feature of the UASs and UGVs is to detect and avoid obstacles while operating autonomously. This thrust explores the possibilities of using low cost off-the-shelf and open sources devices to develop fully autonomous obstacle avoidance UASs and UGVs and train the new generation undergraduate and graduate students. This autonomous air and ground vehicles will extend the opportunity of developing remote sensing sensor platforms for monitoring and sensing environmental changes as well autonomy in the precision agriculture.
Information retrieval approaches for signals of opportunity must contend with many more unknown variables within the measurement scene and the received signal than traditional microwave remote sensing. The received signal is a composite of coherent and incoherent signals. The received signal emanates from quasi-random locations (i.e., non-repeating ground tracks). A combination of effects from vegetation, topography, surface roughness, soil type, and water bodies under bistatic geometry can suppress the soil contribution. Furthermore, non-geophysical factors such as a variation/uncertainty of transmitter power and the receiver antenna pattern corrections can impact the received signals. Empirical models are great for inversion with limited observations but too simplistic to describe such a complex phenomenon. Sophisticated EM models are excellent for describing the physics but need many parameters that do not exist for each observed pixel for retrievals. While not perfect, machine learning offers some benefits for retrieving non-parametric models. In this thrust, we strive to make machine learning models physically sound and constrained by blending them with physical EM models to improve scientific returns. The best example of this research thrust is the “Fusion of Reflected GPS Signals with Multispectral Imagery to Estimate Soil Moisture at Subfield Scale from Small UAS Platforms” via machine learning algorithms.
Agroecosystems are vital to agriculture-based economies. Microwave remote sensing, particularly with small UAS-based L-band radiometry, is an underexplored but promising area in agriculture. In this thrust, we aim to develop a small UAS-based L-band microwave radiometer to retrieve critical information from the earth’s surface, such as soil moisture and vegetation water content. This technology will evaluate the effectiveness of RF sensors in precision agriculture, specifically in assessing water utilization within agroecosystems. By repurposing satellite communication and navigation signals, this approach enables microwave remote sensing at L-band frequencies, distinct from those used in multispectral imaging, to enhance UAS-based agricultural mapping and analytics. Developing a compact L-band microwave radiometer will make satellite-based microwave technology accessible to the general public, especially farmers.
With the growing capabilities of UASs and UGVs, there is an increasing demand for real-time data processing to enhance decision-making and reduce the latency associated with traditional data transmission and analysis. This thrust focuses on the integration of artificial intelligence (AI) and machine learning (ML) technologies directly on UAS platforms to enable real-time environmental monitoring and analysis. By embedding AI algorithms and ML capabilities within UAS, this thrust aims to process sensor data on-the-fly, allowing for immediate interpretation and response. Apply these technologies to real-time applications such as monitoring crop health, assessing soil moisture, detecting environmental hazards, and managing water resources in agroecosystems. This approach is particularly beneficial for applications such as disaster response, precision agriculture, and environmental monitoring, where timely data analysis is critical. This thrust will not only advance the field of UAS and UGV-based remote sensing but also bridge the gap between data collection and actionable insights, making environmental monitoring more efficient and responsive.
Uncrewed Aircraft Systems (UASs) and Uncrewed Ground Vehicles (UGVs),
Applied electromagnetics, microwave, signal processing, and RF Electronics,
UAS-based land remote sensing for precision forestry and disaster response,
GNSS-R bistatic radar, microwave radiometry, and signals of opportunity,
Software-defined radio and RF sensors-based receiver development,
AI and ML in spectrum coexistence, sensing, recycling, and wireless systems,
Robotics, Small UAS-based microwave remote sensing for precision agriculture,