In this article, we insist on the importance and the challenges of the prediction of solar eruptive phenomena including flares, coronal mass ejections (CME), and filament eruptions fully based on the ground-based telescopes. It is true that satellites' data are indispensable for the space weather prediction, but they are vulnerable to the space weather effects. Therefore, the ground-based telescopes can be complementary to them from the viewpoint of space weather prediction. From this view point, one possible new flare prediction method that makes use of H-alpha, red wings, and blue wings images obtained by the SDDI/SMART, the ground-based telescope at Hida Observatory, is presented. And in order to show the possibility for the actual operation based on that method, the recent progress of CHAIN project, the international observation network, is mentioned in terms of their outcomes and capacity buildings.