Neutrino physics is entering a precision era, with experiments targeting goals including CP violation and astrophysical observations. This thesis contributes to three major efforts: the T2K ND280 Upgrade, protoDUNE, and DUNE.

For T2K ND280, it details the construction, development and operation of the Time-of-Flight detector, improving particle identification and background rejection. The system achieved ~175 ps resolution and stable performance over 2023–2025, supported by a newly developed reconstruction framework.

At protoDUNE, a cryogenic beam monitor based on DAMPE silicon tracker technology was developed and deployed, successfully collecting beam data in 2025 and enhancing detector response studies.

For DUNE, a fast supernova neutrino pointing framework was developed using trigger-level data. Combining classical and machine learning algorithms in a controlled fashion, it enables direction estimates within minutes, supporting rapid supernova alerts via SNEWS.

Overall, the work spans from detector development to analysis, advancing the capabilities of next-generation neutrino experiments.