Optimising glass production: 5 levers for more stable campaigns and better delivery performance
Five practical levers to connect demand, campaigns, capacity and shopfloor execution in glass manufacturing – with validated benchmarks and examples.
Table of content
Listen to this article
A glass furnace cannot simply be paused because demand, raw materials or customer priorities change. This is one of the fundamental challenges of glass manufacturing: production needs stability, while markets and supply chains remain dynamic.
In container glass, this tension is particularly visible. According to the European Container Glass Federation FEVE, furnaces operate continuously and typically run for 10 to 15 years before a major rebuild. At the same time, colour campaigns, moulds, available raw materials, stock levels and customer demand all influence what can be produced, where and when. FEVE on continuous container glass production
Key takeaways
- Glass planning needs to absorb variability before it becomes a production problem.
- Data-driven demand planning can significantly reduce forecast error. The key is to combine commercial signals, historical data and operational reality.
- Optimised production planning can reduce changeovers and highlight cost potential, particularly in complex campaign environments.
- The greatest value comes from connecting demand planning, supply planning, scheduling and production rather than managing them separately.
Lever 1
Understand demand earlier – before it changes the campaign
Beverage, food, pharmaceutical, cosmetics and fragrance customers do not follow the same demand patterns. Seasonal peaks, promotions, new products and smaller customer-specific runs all place different demands on planning.
| Customer segment | Typical challenge | Planning focus |
|---|---|---|
| Beverage | Seasonal peaks, promotions and new product introductions | Identify demand earlier and prepare campaigns in time |
| Food | High need for continuous supply | Protect delivery performance with appropriate stock levels |
| Pharmaceutical | High quality and traceability requirements | Connect production, quality and documentation reliably |
| Fragrance & cosmetics | Smaller volumes, individual designs and demanding quality requirements | Plan smaller campaigns and specific requirements efficiently |
Robust demand planning therefore needs to go beyond extrapolating historical volumes. Statistical forecasts, sales knowledge, promotions and current demand developments need to come together in one process.
A McKinsey analysis from the consumer goods sector illustrates the potential of more advanced methods: it describes reductions in forecast error of 30 to 50 per cent for advanced machine-learning-based demand planning. This is not a guaranteed result for the glass industry, but it demonstrates why forecast quality can be an important operational lever. McKinsey: Supply Chain 4.0 in Consumer Goods
For glass manufacturers, however, accuracy alone is not enough. What matters is whether changes are detected early enough to adjust stock, campaigns and capacity while there is still room to act.
Lever 2
Optimise campaigns instead of reacting to the next bottleneck
Colour and product changeovers are among the key constraints in glass planning. Every changeover consumes time and capacity. At the same time, campaign structures that are too rigid can leave stock and customer demand out of alignment.
A 2026 scientific case study on the production planning of glass jars and vacuum flasks shows the potential impact of optimised sequencing. In a real quarterly planning instance, the optimisation approaches examined reduced the number of setups by 26.3 per cent compared with the existing factory plan. Modelled total production costs were 3.4 to 4.5 per cent lower, depending on the optimisation method used. Study on optimised production planning in glass manufacturing
These figures are not a general performance promise. They do, however, show why scheduling in glass manufacturing is about much more than moving orders around a Gantt chart: sequence, changeovers, capacity, demand and stock all affect one another.
Factory Scheduling helps planners consider these constraints together and compare alternative production plans before committing to a decision.
Lever 3
Plan Hot End and Cold End as one process
An optimised Hot End plan is not enough if downstream processes cannot keep pace. Forming, further processing, quality inspection, packing or outsourced decoration can introduce their own constraints and capacity limits.
Connecting Hot End and Cold End is therefore an important part of effective glass planning. Production batches and downstream sub-batches need to be coordinated without creating local optimisations that simply move the bottleneck elsewhere.
Multi-scenario planning can bring Hot End and Cold End processes into the same planning logic. Colour campaigns, furnace maintenance and other production constraints can then be considered when comparing alternative planning options.
Lever 4
Compare plants and capacity through scenarios, not coordination loops
With multiple plants, the number of options increases – and so does the complexity. Can a reference be produced at another site? Is sufficient capacity available there? How would the move affect other customers? And what would it mean for stock and logistics?
Questions like these are difficult to answer with isolated spreadsheets. Scenario planning makes it possible to compare alternatives using a common data basis before a decision is implemented.
Verallia provides one example of how this can work in practice. The international container glass manufacturer uses sedApta to connect S&OP, Master Production Scheduling and Shop Floor Scheduling. Published project outcomes include improved forecast quality, more stable production programmes, greater simulation capabilities, stronger interaction between MPS and scheduling, and faster decision-making.
The key point is not one individual algorithm. It is that planning levels build on one another and work from consistent assumptions.
Lever 5
Connect planning and execution
Even a strong production plan quickly loses value if the actual situation on the shopfloor only feeds back into planning with a delay.
This is where scheduling and MES close the gap between planning and execution. Order progress, machine status, scrap and other production data provide the basis for comparing the plan with operational reality. OEE is not an end in itself, but a metric that helps make losses and improvement opportunities systematically visible.
One example is Lalique. The crystal manufacturer uses MES, scheduling and maintenance management. The project provides targeted and reliable real-time production information and has reduced administrative effort associated with data entry and management, while supporting stronger production control and productivity.
For glass manufacturers, this creates an important closed loop: plan, execute, identify deviations and reassess.
Which KPIs show whether glass planning is working?
Good planning should not be judged by how detailed the plan looks. What matters is the effect it has on operations.
| Planning area | Relevant KPIs | What they show |
|---|---|---|
| Demand | Forecast error, forecast accuracy by segment | How reliably demand is identified early |
| Campaigns & scheduling | Number of setups, duration of colour and product changeovers, campaign adherence | How stable and efficient production is |
| Inventory | Stock levels, days of inventory, working capital | How well production rhythm and demand are balanced |
| Customer service | OTIF, product availability, unplanned urgent orders | How reliably planning translates into customer service |
| Production | OEE, scrap, production losses | Where operational losses occur |
| Planning process | Planning effort, replanning frequency, response time | How quickly planning can respond to change |
These KPIs should not be optimised in isolation. Lower inventory, for example, is not an improvement if delivery performance or campaign stability suffers. Equally, high furnace utilisation adds little value if the products customers actually need are unavailable.
The real leverage lies between the systems
Many glass manufacturers already have ERP systems, planning software, spreadsheets, shopfloor applications and large volumes of production data. The challenge is often not a lack of data, but the connection between decisions.
Most planning software, for example, was developed for industries such as automotive or pharmaceuticals – manufacturing environments that operate very differently from container glass. If these systems fail to take account of industry-specific constraints such as mould changes, maximum tonnage or multi-stage decoration processes, planners often do the only sensible thing: they step outside the standard system process and return to Excel spreadsheets or their own calculations.
For Elisa Industriq, effective glass planning therefore means validating demand collaboratively, aligning stock and capacity, comparing alternative scenarios, generating realistic production sequences and feeding actual execution back into the next planning decision.
The sedApta suite brings together capabilities for Demand Management, Resource & Supply Planning, Factory Scheduling and MES.
Conclusion: glass production does not become more stable through more planning, but through better decisions
The specific constraints of glass manufacturing cannot simply be digitalised away. Furnaces run continuously. Campaigns require stability. Raw materials need to be available. Customer demand will still change.
The difference lies in how early those changes become visible and how quickly their impact can be assessed.
By connecting demand, inventory, capacity, campaigns and the shopfloor, glass manufacturers create a stronger basis for decision-making – not for producing the perfect plan, but for making better decisions when reality diverges from it.