Large areas of astrophysics, such as precision cosmology, have benefited greatly from large maps and datasets, yielded by telescopes of ever-increasing number and ability. However, due to the unique challenges posed by submillimeter polarimetry, the study of molecular cloud dynamics and star formation remain stunted. Previously, polarimetry data was limited to a few vectors on only the brightest areas of molecular clouds. This made drawing statistically-driven conclusions a daunting task. However, the successful flight of the Balloon-born Large Aperture Submillimeter Telescope for Polarimetry (BLASTPol) generated maps with thousands of independent polarization measurements of molecular clouds, and ushered in a new era of empirical modeling of molecular cloud dynamics. Now that the potential benefits from large-scale maps of magnetic fields in molecular clouds had been identified, a successor that would truly unlock the secrets must be born. The Next Generation Balloon-borne Large Aperture Submillimeter Telescope (BLAST-TNG), the successor to BLASTPol, has the ability to make larger and more detailed maps of magnetic fields in molecular clouds. It will push the field of star formation into a statistics-driven, empirical realm. With these large, detailed datasets, astronomers will be able to find new relationships between the dust dynamics and the magnetic fields. The field will surge to a new level of understanding. One of the key enabling technologies of BLAST-TNG is its three arrays of polarization-sensitive Microwave Kinetic Inductance Detectors (MKIDs). MKIDs are superconducting RLC circuits with a resonant frequency that shifts proportionally to the amount of incident radiation. The key feature of MKIDs is that thousands of detectors, each with their own unique resonant frequency, can be coupled to the same readout line. This technology will be able to drive the production of large-scale monolithic arrays, containing tens or hundreds of thousands of detectors, resulting in an ever-increasing rate of scientific progress. The current limiting factor that determines how many MKIDs can be placed on the same readout line is the bandwidth and processing limitations of the readout hardware. BLAST-TNG has pushed this technology forward by implementing the first Reconfigurable Open-Architecture Computing Hardware (ROACH2) based readout system. This has significantly raised the processing abilities of the MKID readout electronics, enabling over 1000 MKIDs to be read out on a single line. It is also the first ever ROACH (1 or 2) based system to ever be flown on a long duration balloon (LDB) payload. This thesis documents the first-ever deployment of MKIDs on a balloon payload. This is a significant technological step towards an MKID-based satellite payload. This thesis overviews the balloon payload, details the underlying detector physics, catalogs the detector and full-scale array development, and ends with the room-temperature readout electronics.