To naively smell as no robot has smelt before

Abstract—This paper ,presents a new ,intelligent odor localization strategy, which enables a robot to locate the source of anodor in a cluttered indoor environment. Traditionally, work inthis area has focused on open ,areas free of obstacles and having no walls ,or possessing walls without openings. Existing solutions predominantly use reactive algorithms to navigate along the entire length of the odor plume to the source. Not only isthis slow, but in a cluttered indoor environment it may not be possible. In a constrained environment, airflows tend to circulate insectors and well-defined plumes that lead upwind to the odor source do not ,exist. We have ,developed a sense-map-plan-act style control strategy to model ,the airflow in the ,environment using naïve physics, then use the model to reason about odor dispersal, move to key positions gathering information, and make aprediction of the most likely location for an odor ,source. The control strategy has located the odor source for a variety of room configurations. This paper ,describes details of the ,control strategy, practical experiments, and results. Keywords— odor localization, naïve physics, airflow modeling