The protostellar phase, an early phase of star formation, has recently gained the spotlight as a phase for planet formation. On the other hand, the place of planet formation, “disk”, is difficult to identify in the protostellar phase because disks are embedded in envelopes in this phase. Identifying a disk also tells us the most important parameter of the protostar, central stellar mass M*. Previous works reported observational identification of disks in protostellar systems. The method is, however, different from work to work. To verify the consistency among different methods, I applied representative methods to synthetic observations using a magnetohydrodynamic simulation of protostellar evolution. This test demonstrates that a method using position-velocity diagrams can estimate the disk size and the central stellar mass accurately, particularly when M* >~ 0.2 Msun. Our group has been using this method with SMA and ALMA observations almost for a decade and are continuing to develop it. Even after a disk is identified in a protostar, the system provides questions related to the disk and planet formation: e.g., how the disk obtains mass or how dust grains grow in the disk. I introduce a case study of the protostar TMC-1A to investigate these questions, based on dust continuum, spectral line, and dust polarization observations using the SMA and ALMA. In addition to these studies I accomplished in KASI, I would also like to review my activities in KASI during the ~3-yr term.