Home / Solutions / Production and maintenance
Planning task
Production scheduling decides which order runs when on which machine, and who operates it. Optimized means: due dates are kept, set-up times and colour changes go down, and maintenance and energy prices flow into the same plan.
Rules in the model
From practice
Project example
Allnex: market leader in industrial coating resins, 34 production sites worldwide
Use cases
Every case plays in your browser: the plan in time-lapse, with KPIs and the comparison to the usual way of planning.
09:20CNC mills, lathes, welding and assembly; component orders feed assembly, AGVs move the material, workers in shifts.
10:00Laser cutting, bending, robot welding and a powder coating line that runs in colour batches.
04:00Wash plant, inspection pits, drop pit and wheel lathe; one shunting crew moves the trains through the switch ladder.
11:00Induction furnaces, heat treatment, e-trucks, PV and a battery planned together against day-ahead prices and the 15-minute peak.
Weeks28 real turbines, three crews, one mobile crane and a weather front. Services run in calm hours, faults are fixed quickly.
WeeksA simulated day with 863 AI jobs on 9,728 GPUs in five data centres. The OptaZEN meta-scheduler plans 24 hours ahead in front of the existing scheduler: which job runs when and where, when data moves and when renting extra GPUs pays off.
WeeksNine printers for titanium, cobalt chrome, steel, PA12 and resin, 20 post-processing stations, two shifts. Which lot goes on which build plate and printer, and in which order, so surgery dates hold, also after a failed build.
01:45About 120 dough batches, seven kneaders, four make-up lines, eleven ovens and 15 tours that leave between 01:45 and 05:30. Dough cannot wait, so the order of the doughs decides the night.
Weeks82 work orders for two weeks, nine own technicians and four external companies. Maintenance uses the gaps of the production plan, and the bottlenecks stop as little as possible.
05:00400 utility vehicles in many variants, 32 stations at a takt of 72 seconds, a paint shop that wants long colour blocks and suppliers who deliver just in sequence.
06:00Four bays, two aircraft in a heavy check and three overnight checks every night. Every aircraft has to be released before its next flight.
Approach
Stage 1Clarify
DecisionIs it worth it?
Stage 2Build
DecisionDoes the pilot deliver what it promised?
Stage 3Run
CycleWhen operations change, we adapt the rules and the plan.
Questions
No. Orders come from your ERP or MES, and the optimized plan goes back there.
A rush order is planned in, and the plan shows which other orders move because of it.
We check that in the first call and in a pilot with your data. The use cases show how the model groups orders with similar set-ups.
The first step
In a free first call we work out which next step is really worth it.
