Confident promises, on a moving due date

The situation

A steel manufacturer making formed and finished products for the construction industry across five plants. Construction is a hard environment to promise into. Lead times are very short, and customers move their due dates constantly as conditions on site change, so the date agreed last week is not the date that matters this week. Most of that uncertainty comes from outside the business entirely. Inside, the process is machine-intensive forming and finishing, where every changeover between products costs time on the very machines that constrain output. And the plants are a network, not islands: some carry the sales relationship and are supplied by others, so a promise made in one place depends on capacity in another.

The recurring decisions are two, tightly coupled: what to run, and in what order, on constrained machines; and what to promise a customer, and when. Both have to be made fast, repeatedly, against demand that will not hold still.

The work

Kinetica put a finite scheduling approach at the centre, one that combines production orders to balance two things that usually fight: the fast response and short lead times the market demands, and the setup reduction the machines need to stay productive. It stays flexible by design, so a schedule can be reworked as demand and due dates move rather than broken by them. Part of the work was simply defining standards where none had existed, so the same decisions were made the same way across the network. The payoff was visibility in two directions: supply chain could see real capacity, and sales could see what the plants could actually deliver.

The result

With that visibility, sales could accept orders and make promises they could keep, capable-to-promise grounded in real capacity rather than optimism, even with due dates shifting underneath them. The finite scheduling balanced responsiveness against changeover cost across a broad product range and five interdependent plants, and gave the business a common standard where it had been running on local habit.

The capability was not a one-off schedule. It was a way of scheduling and promising that held up as demand kept changing, across plants that depend on each other.