Torpedo cars planned 48 hours ahead and replanned in under a second

Up to 96%

of the torpedo car plan executed as planned in a single month; 82% on average over six months of operation

48 h

of lead time in planning each torpedo car action

< 1 s

to replan the fleet, in tests with different fleet sizes

Synchronized maintenance

to avoid several cars being out of service at the same time

Torpedo cars carry liquid hot metal from the blast furnace to the steel shop. When the fleet is not in the right place, the mill loses at both ends:

• Too few cars. When the steel shop calls for a lot of hot metal and there is no empty car at the blast furnace, production drops on both sides.
• Too many cars. When the steel shop runs slowly, full cars pile up. After 18 hours of waiting, the hot metal is tipped at the emergency yard: lost raw material, energy to reprocess it, and environmental impact.
• Cold or idle car. Hot metal that cools needs reheating. Each car has a limited service life before a relining that can take up to 20 days, and it needs periodic cleaning.

With so many constraints at once, the fleet was used inefficiently and created bottlenecks in the process.

With Enacom’s Planning Digital Twin, every fleet trip is now planned from what is happening at the blast furnace and in the steel shop:

1. Allocates each car

   considering position, remaining load, queues at the converters, blast furnace output, cleaning, and maintenance, and plans its actions up to 48 hours ahead.

2. Replans in real time

   from the status of each car, in seconds, prioritizing cars with more remaining service life and keeping full cars moving to preserve the hot metal temperature.

3. Synchronizes maintenance and alerts before problems occur

   it avoids simultaneous relinings, flags the end of service life for the operator’s decision, and warns of blast furnace stoppage risk, unmet demand, and inconsistent data.

When the team takes a different path, the system asks for a justification, which makes it clear who decided what.

The operation can now see the fleet up to 48 hours ahead and replan automatically at every change. In tests with different blast furnace output levels and fleet sizes, a full replan took less than a second, fast enough to keep up with the operation in real time.

In operation, the plan generated by the twin was followed as proposed in up to 96% of the actions in a month, with an average of 82% over six months: the team now executes the plan, not just consults it.

The project records the following results:

• Higher fleet availability, with less downtime and no simultaneous maintenance.
• Less hot metal wasted in emergency dumps, and hot metal delivered at the right temperature.
• Longer car service life, less frequent maintenance, and fewer cars needed for the operation.
• Traceable decisions and problems anticipated for the coming shifts.

• More hot metal put to use. Fewer cars waiting idle and fewer emergency dumps.
• A fleet that delivers more. Maintenance and cleaning spread over time, with priority for the cars with the most service life.
• The coming shifts in sight. Risks such as a blast furnace stoppage for lack of a car show up before they happen.
• Decisions made with the team and on the record, in a solution that integrates with the mill’s systems.

• Your mill carries hot metal from the blast furnace to the steel shop in torpedo cars.
• Emergency dumps, queues at the converters, or cold cars are recurring.
• Fleet allocation and car maintenance are still decided shift by shift, by hand.

We start from your data: we customize the twin, validate the result in scenarios and a pilot, and only then scale. The results in this case reflect the conditions of that mill.