The digital model «Ingot Casting» has been developed to optimize the process of continuous casting of aluminum ingots, aimed at improving product quality and process safety. The model solves the key problems of traditional methods: high resource intensity of experiments, difficulty in predicting defects (cracks, shrinkage cavities) and the risks of accidents associated with a breakthrough of liquid metal (~700 ℃) with insufficient thickness of the crust zone. Practical verification was carried out for the AlMg5 (1580) alloy. The model identified dangerous conditions: at a casting speed of 80 mm/min and a water flow rate of 250 l/min, the thickness of the crust zone at the outlet of the crystallizer reached 0 mm, and the Niyama criterion (1.28) indicated a risk of porosity. The model can be integrated into an automated process control system via industrial protocols, providing prompt recommendations for adjusting parameters. This reduces the risk of defects and eliminates accidents associated with metal breakthroughs. Further development includes predictive control algorithms and adaptation to new alloys.