AI-powered multimodal precision sensing system to track tick phenotypic responses in natural and simulated environments - The long-term goal of our team (Drs. Hill, Murgia and Kaur) is to revolutionize control of vector-borne diseases through innovations in autonomous, AI-driven robotics systems for precision pest management. On this two-year R21, we propose the development of a transformative autonomous robotic system for precise, real-time detection and targeted treatment of the Lyme disease tick, Ixodes scapularis in occluded natural environments. We will leverage our combined expertise in tick biology, AI, sensing, robotics and product development to introduce two novel technologies: (1) a mmWave-sensing system customized for detection and tracking of multiple nymphs and adult ticks in occluded environments and (2) AI-based algorithms to drive the autonomous, large-scale sampling of multiple land types, and subsequent detection of tick habitat and tick hotspots by quadruped robots (or alternatively, drones). In Aim 1, we will develop an AI- powered multimodal sensing-based system taking advantage of the Purdue Phenocosm, a versatile test arena for detailed assessment of tick behaviors over weeks-long study periods and involving various substrate types to mimic natural tick habitats. This work will combine multi-modal mmWave and traditional camera technology, together with neural network-driven tracking algorithms. Through this work, we expect to generate the first system capable of continuously monitoring multiple ticks in complex, occluded environments, representing a paradigm shift for surveillance. In Aim 2, we will develop sophisticated path planning, navigation, and control algorithms needed to support autonomous quadruped robot movement in outdoor settings, including a range of habitats such as deciduous forest/grass ecotones that typically support I. scapularis populations. The quadruped robot will have capability to (1) navigate in and adapt movements for uneven and complex terrains, (2) optimize movement for energy efficiency, (3) collect large environmental data sets and (4) detect ticks concealed by vegetation, leaf litter and other matter. The sensing and robotics technologies developed on this project will lay the foundation for future field studies to validate protypes for area-wide tick surveillance and targeted control. Once validated and fully developed, this autonomous system will greatly expand U.S. public health surveillance capacity, substantially reducing human-tick encounters, the incidence and burden of tick-borne diseases, pesticide use, healthcare and pest management costs, paving the way for a new era in fully remote pest and disease management. 1
