The stars evolving through the asymptotic giant branch (AGB) are generally regarded as highly efficient dust manufactures, owing to the thermodynamic properties of their wind, which prove extremely favourable to the condensation process of gas molecules into solid grains. In this review I will describe the dust and mineralogy of the dust  formed in the surroundings of this class of stars, outlining the role of mass and metallicity, and the importance of these studies for the characterization of evolved stellar populations in galaxies. The contribution from the analysis of the spectral energy distribution of post-AGB stars towards a better understanding of the dust formation process by AGB stars will be also commented on.

On their way from the main sequence to the final supernova explosion, massive stars lose a substantial fraction of their mass through line-driven winds. Recent decades have witnessed significant advancements in both observational and theoretical studies of these winds that sail on starlight. The advancements in our understanding of radiative driving lead to progressively more accurate estimates of mass-loss rates from massive stars. In this talk, we will outline the key ingredients necessary for reliable predictions of mass-loss rates from numerical simulations, and demonstrate how state-of-the-art theoretical mass-loss rate estimates compare with observational results.

High resolution, hydrodynamic galaxy simulations can be used to investigate the inherent variation of dark matter around the Solar Circle of a Milky Way-type galaxy. These simulations self consistently include both the baryonic back-reaction as well as assembly history of substructures, all of which may have lasting impacts on the dark matter’s spatial and velocity distributions, creating `gusts’ of dark matter wind around the Solar Circle, potentially complicating interpretations of direct detection experiments on Earth. Direct detection is a key experimental goal to advance the microscopic understanding of the dark matter that fills the Universe. We investigate how dark matter substructure, simulated in halos analogous to our own Milky Way, impacts the shape, summary statistics, and interpretation of results from terrestrial dark matter direct detectors.

Implementing a new numerical integration technique, our work generates bespoke predictions for terrestrial underground detection, finding large uncertainties arising in the expected signals of direct detection experiments. Having developed a realistic end-to-end pipeline for studying these effects, we discuss the implications of these astrophysical variations in the dark matter distribution of the solar neighbourhood on current and future particle physics searches for dark matter.