Autonomous Swarm-enabled Vegetation Control

Weeds and invasive vegetation pose a great risk on military bases as they can be a source of foreign object debris (FOD) and may attract wildlife, introducing bird/wildlife aircraft strike hazard (BASH). Air Force Instruction (AFI) 91-212 outlines the risks and dangers that weeds and invasive vegetation can have in Aircraft Movement Areas and provides specific guidance on ground management to mitigate those risks. Existing abatement solutions for weed and invasive vegetation control are labor intensive and utilize harsh chemicals that do not promote agricultural sustainable practices and exposes airmen to toxic herbicides. Additionally, current practices are highly susceptible to disruptions from labor shortages and supply chain of herbicides due to geo-political events, presenting unsustainable supply. This presents a need to find and establish sustainable vegetation control measures that can alleviate the dependency on human-labor and toxic chemical. Novel Engineering proposes to utilize a swarm-enabled autonomous robotic solution called RISE (Rapid Integration Swarm Ecosystem) to conduct ground management operations for weed and vegetation control. Our objective is to adapt and implement the RISE system to schedule and coordinate the operations of autonomous UGVs for agricultural ground management using modern AI/ML computer vision technology and concept vehicle prototype vehicles that are COTS based with custom designed fitments: 1) a precision CO2 focused laser and 2) precision manipulated string trimmer payload. We aim to build a functioning system that can be experimented with and tested in relevant environments to develop a more reliable solution to weed and vegetation control. through Alternative Piloting modes inside RISE, these specialized robots will be able to perform vegetation control autonomously and also learn from human operators on necessary behaviors and operations in complex situations through remote piloting mode, that enables humans to remotely control all aspects of the UGV and its payloads. This will also provide a pipeline data to feed into the onboard AI to learn these complex operation behaviors and better perform them autonomously in the future. RISE also enables intelligent onboard computer vision and edge processing capable of detecting and identifying undesired vegetation along with specified plant anatomy attack points using the Segment Anything Model. The target position will be used by the onboard vegetation control measures to precisely eliminate the undesired weed and invasive vegetation. RISE will be able to provide scaling of this capability to a swarm of vegetation control UGVs using its swarming architecture and UI to coordinate and schedule operations during optimal times on base. Vegetation control becomes more routine, no longer relies on a large workforce, eliminates the dependency on harsh toxic chemicals, and creates sustainable agricultural practices that are ripe for widespread adoption.