The broader impact of this SBIR Phase I project is to improve habitat for native plants and wildlife by removing invasive plants with a mobile robot. Worldwide invasion by non-native plant species is a primary concern in forest ecosystem health and biodiversity. Within the highly fragmented landscape of the Connecticut forest, two invasive plants significantly contribute to the degradation of habitat: Japanese barberry and multiflora rose. Current methods to eradicate them are time-consuming, expensive, and ultimately ineffective. This project presents a novel method of detecting and removing Japanese barberry and multiflora rose. Deep learning technology enables additional use with other invasive species, increasing the system's value throughout natural resource management. Additionally, once programmed, the robots will be easily operated, making them usable for a variety of personnel. This SBIR phase I project will advance current robot prototypes being tested for weed detection and removal. There is currently no mobile robot designed and programmed to remove understory invasive shrub species in a deciduous forest ecosystem. The technical challenges that will be addressed in building a feasible robot prototype include navigating over the unstructured terrain of the forest floor, developing a cutting attachment for the robotic arm, and creating a hybrid soft-rigid platform to withstand forest floor hazards while averting tree seedling damage. The proposed system will operate as follows: (1) an unpiloted aerial system (UAS, i.e., drone) flies over the canopy to capture images of the forest understory; (2) images are then labeled and converted to species location maps; and (3) utilizing these maps, a semi-autonomous mobile robot navigates to the invasive species locations for removal. The robots will be programmed for semi-automated missions monitored by an operator nearby at a safe distance to ensure worker safety.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
