Unmanned aerial vehicles are rapidly gaining popularity in many environmental monitoring tasks. A prerequisite for their autonomous operation is the ability to perform efficient and accurate mapping online, given limited on-board resources constraining operation time and computational capacity. To address this, we present an online adaptive-resolution approach for field mapping based on Gaussian Process fusion, a strategy in which Bayesian fusion is applied to update a Gaussian Process prior map. A key aspect of our approach is an integral kernel encoding spatial correlation over the areas of grid cells. This enables efficient information compression in uninteresting areas to achieve a compact map representation while maintaining spatial correlations in a theoretically sound fashion. We evaluate the performance of our approach on both synthetic and real-world data. Results show that our method is more efficient in terms of mapping time and memory consumption without compromising on map quality. Further, we integrate our mapping strategy into an adaptive path planning framework to show that it facilitates information gathering efficiency in online settings.

In robotic applications, a key requirement for safe and efficient motion planning is the ability to map obstacle-free space in unknown, cluttered 3D environments. However, commodity-grade RGB-D cameras commonly used for sensing fail to register valid depth values on shiny, glossy, bright, or distant surfaces, leading to missing data in the map. To address this issue, we propose a framework leveraging probabilistic depth completion as an additional input for spatial mapping. We introduce a deep learning architecture providing uncertainty estimates for the depth completion of RGB-D images. Our pipeline exploits the inferred missing depth values and depth uncertainty to complement raw depth images and improve the speed and quality of free space mapping. Evaluations on synthetic data show that our approach maps significantly more correct free space with relatively low error when compared against using raw data alone in different indoor environments; thereby producing more complete maps that can be directly used for robotic navigation tasks. The performance of our framework is validated using real-world data.

Sulfur is of fundamental importance in a wide variety of phenomena such as life on Earth. This element is one of the most abundant elements in space S/H - 1.3×10^-5, but sulfuratted molecules are not as abundant as expected, thus a better understanding of sulfur chemistry is needed. We study and model the abundance of H₂S in two prototypical dark clouds, TMC1 and Barnard 1, to shed light on the physical and chemical processes involved in H₂S creation and destruction. Our observations are consistent with a PDR model in which H₂S is formed in grain mantles and released to gas phase via photodesorption. We cannot discard the contribution of other desorption processes, such as chemical desorption and/or grain-grain collisions, to enhance the H₂S abundance.