Brown dwarfs (BDs) lie in the substellar mass regime, and are a bridge between stars and planets, thus providing a unique window into the unknowns of their formation processes. The dominant mechanism behind the formation of BDs is still not fully constrained. In fact, BD formation is expected to be affected by the environment in which they are born, in particular, an environment with high stellar densities and/or large number of massive stars would increase the efficiency of BD production compared to stars. In order to test these hypotheses, we are studying the low-mass population of three massive young clusters with extreme environmental properties compared with nearby star-forming regions. One of these clusters is NGC 2244 (d=1.5 kpc), which hosts a large number of OB stars and presents a low stellar density. We have built a robust sample of cluster members using deep photometry, astrometry and multi-object spectroscopy (VIMOS and KMOS/VLT), resulting in the first spectroscopically confirmed BDs beyond 1 kpc. In this talk, I will present our newly developed method for spectroscopic analysis of cool dwarf spectra in the NIR, implementing machine learning models to efficiently separate young members from field contaminants, that will be of special interest for upcoming multi-object facilities such as NIRSPEC and NIRISS/JWST and MOONS/VLT. Furthermore, I will present our results on the BD population in NGC 2244 and compare them with the other massive clusters in our sample, as well as with nearby star-forming regions.