Glass Manufacturing Scheduling Around the Tank Furnace

Glass manufacturing scheduling has to start from a hard fact: the tank furnace runs continuously at a fixed rate and cannot be sped up to chase a busy day. EDGEBIC by User Solutions models the furnace as a continuous process and plans forming, annealing, and inspection against the throughput the melt can actually feed, so the plan never promises more finished ware than the furnace can supply. For a glass plant where the furnace paces everything, that constraint-first approach is what makes a schedule trustworthy.

The Furnace Is a Continuous-Rate Constraint

A batch machine can be double-shifted to catch up. A glass tank furnace cannot. It runs at a fixed pull rate around the clock, and everything downstream can only process what the melt feeds it. A scheduler that plans forming against forming's own theoretical capacity will promise finished ware the furnace cannot supply, and the promise breaks the moment the melt cannot keep up.

Model the furnace as a continuous process, and the scheduler treats it as steady flow at its rate. Forming, annealing, and inspection are then planned against what the furnace feeds, so the plan paces the whole line to the melt. The furnace becomes the pacing constraint the rest of the schedule respects. The continuous model is documented in continuous process work centers, and the same approach is used in Li-ion battery continuous process.

Planning From the Constraint

When one resource paces the plant, the schedule has to be built around it. If forming could in principle run faster than the furnace supplies, the plan still paces forming to the melt, so it never promises ware the furnace cannot support. That is the practical meaning of scheduling from the constraint, and it is why the furnace has to be the pacing resource rather than an afterthought.

It also changes what an improvement looks like. When you want more throughput, the plan shows the furnace as the constraint, which tells you that adding a forming machine will not help until the melt rate changes. That is the kind of answer production bottleneck identification exists to give, and a constraint-first schedule surfaces it instead of hiding it behind an average.

Several Continuous Stages, Then Batch Finishing

Glass does not stop being continuous at the furnace. The annealing lehr also runs continuously, so it is modeled as a continuous process too, planning as steady flow at its own rate. A routing can carry the furnace and the lehr as consecutive continuous stages, feeding batch operations like inspection, decoration, and packing downstream. Fabricators working the far end of that chain, cutting and tempering purchased flat glass, face a different pace-setter: see scheduling architectural glass fabrication around the tempering furnace.

Each stage plans the way it physically runs. The schedule reflects a continuous melt and anneal feeding discrete finishing rather than forcing one model across all of them, which is finite capacity done at the level of the individual machine.

Changeovers That Happen While the Furnace Runs

The furnace never stops, but the forming line changes ware type, and that changeover is real time. The forming work center can carry a per-machine changeover matrix, so a ware-to-ware or color-to-color transition is priced by the pair rather than by a single flat number. Grouping similar ware to minimize those changeovers is then a sequencing decision the matrix makes visible, without ever pretending the melt paused. The mechanism is the same sequence-dependent setup used across mixed-product lines, applied to the one stage of a glass plant that actually changes over.

Keeping the Plan Honest From the Floor

A continuous plan drifts when the pull rate or a downstream line runs off its assumed rate. Operators log actual production and start and end on a shop-floor kiosk, so a stage running below rate is visible in the schedule rather than hidden until the finished-ware shortfall appears. On a reschedule, completed production with recorded actuals is never re-planned, so only the remaining work re-times from where the run actually stands. The workflow is covered in shop-floor actuals tracking.

Before committing a large order against a fixed melt rate, you can test whether the plant holds the promise by running a quote simulation, which uses the same continuous and constraint logic as a live plan without changing anything.

Constraint-first, continuous scheduling has a long track record in the User Solutions and RMDB lineage, which includes finite-capacity work across heavy process and metals industries, including a GE railcar operation that moved from roughly 30 percent to 90 percent on-time delivery. The same discipline that paced that flow to its real constraint is what plans your furnace here.

For the fundamentals, see continuous flow manufacturing, and for a related continuous sector, paper converting continuous scheduling. For precision glass work gated by batch equipment instead of a melt rate, see scheduling optical and lens manufacturing around coating chambers. The industry fit guide maps the rest, and EDGEBIC is the product hub.

Ready to plan around your furnace? Contact US for a demo and bring your melt rate and downstream stage rates.