Input Planning.
Conventional charge planning focuses on scrap chemistry, but optimal charging is also driven by raw-material cost, energy input and achievable yield.  A smart charge-planning module generates a ranked list of recommendations from real-time chemistry, current raw material costs, energy input parameters, and yield targets. The shift supervisor reviews the recommendations and confirms a data-driven decision within minutes, reducing raw material cost per ton and 
improving first-cast quality.

Automation of the Charging Process.
Today, the decision when to reload a furnace is often a judgement call made by opening the door and visually assessing the melt. On a good day, this works. But timing errors, recharging too early or too late, create excess dross, waste energy, and reduce throughput, with measurable variation between day and night shifts. A camera system installed inside the furnace chamber continuously analyzes the melt state and combined with other process parameters like gas, power consumption and furnace temperatures the optimal recharge moment is automatically determined. The subjectivity is removed entirely. Dross formation is reduced, metal yield increases, performance becomes consistent regardless of who is operating the furnace.

Automatic Dosing of Alloying Components. 
Manual alloying addition is a familiar scene: an operator approaches the furnace with bagged material, tosses it into the melt, and waits. The precision of this process depends entirely on the individual, their calculation accuracy, their consistency, and their proximity to the hazard zone. Overshooting is common. Repeatability is poor. Automated dosing systems use calibrated storage containers with precision discharge mechanisms to add Si, Mn, Cu, Fe, and other elements via the charging chute alongside loaded scrap. An electromagnetic stirrer at the furnace bottom ensures rapid dissolution and homogeneous distribution. In furnaces with an open side well, alloying materials can alternatively be added directly to the liquid metal, provided a drying or preheating system prevents moisture introduction. Alloy overshooting is reduced, composition consistency improves, and operators are no longer required near the melt for this task.

Fine-Tuning of alloy composition. 
Magnesium is highly reactive at casting temperatures. Even a well-prepared melt can arrive at the holding furnace with Mg levels below specification due to oxidation losses during launder transfer. Such silent yield loss forces costly rework including additional alloying cycles. Magnesium-ingot submergence devices installed above the launder perform precise, automated Mg addition during transfer, compensating for oxidation losses in real time. Target composition is achieved in a single pass, eliminating correction loops and improving production efficiency from the very first cast.

Automation of the Dross Skimming Process.  
Dross skimming is one of the most physically demanding and hazardous tasks in the casthouse. An operator is controlling a long-wearing skimming rake across the surface of the melt, working in intense heat, often with limited visibility through fume and glare. Mechanic impacts on furnace wall refractory, or the furnace sill are not unusual. The quality of skimming varies from person to person and shift to shift, directly affecting metal recovery and furnace cleanliness. Automated skimming machines equipped with onboard cameras execute the task without human presence at the furnace door, with an operator cabin available for manual override. A supervised learning period trains the system on the specific furnace geometry and dross behavior, afterwards it operates autonomously with consistent technique. Periodic manual furnace-wall and bottom cleaning remains, but the high-frequency, high-exposure skimming task is fully removed from the manual workflow.