“The model showed us exactly which tanks were causing bottlenecks—and the AI-optimised schedule eliminated 80% of our labour excess hours.”
— Production Manager
The Challenge
A vinegar manufacturer was struggling to get full value from their production assets. The operation spanned milling, mashing, fermentation, and bottling—each stage with its own constraints and interdependencies. Storage tanks sat idle while others overflowed. The bottling hall couldn’t maintain consistent throughput. Acetation times varied without clear understanding of why. And shift patterns had evolved around problems rather than optimal production flow.
Management knew there was untapped capacity in the existing facility, but couldn’t pinpoint where the constraints actually were. Was it fermentation? Tank availability? Bottling line speed? Without visibility into how each stage affected the others, scheduling remained reactive and inefficient—with excess labour hours, energy waste, and throughput below potential.
The Solution
We built an interactive model of the complete production process—from milling and mashing through fermentation to the bottling hall. The model tracked tank levels, utilisation, and flow rates at every stage, then used AI to optimise production scheduling and shift patterns.
Tank Utilisation Analysis
The model tracked tank levels across the facility, revealing which tanks were genuinely required and which were masking scheduling problems. By modelling fill rates, hold times, and draw-down patterns, we identified that some tanks were bottlenecks while others were underutilised. This insight allowed us to reduce the number of storage tanks while actually increasing overall utilisation to 87%.
Constraint Identification
The model mapped constraints at each stage of production, pinpointing exactly which tanks and processes were limiting throughput. It revealed opportunities to reduce acetation time that had been invisible in day-to-day operations. With constraints clearly identified, we could address the real bottlenecks rather than working around symptoms.
AI-Optimised Scheduling
We applied AI to optimise production scheduling across the entire operation. The algorithm balanced milling and mashing rates against fermentation capacity and bottling hall demand, generating schedules that minimised idle time and eliminated labour excess hours. Shift patterns were redesigned around optimal production flow rather than historical practice.
Bottling Hall Optimisation
By relieving upstream constraints, we improved throughput rates in the bottling hall. The model showed how fermentation timing and tank availability directly affected bottling line performance—and how synchronising these stages could unlock capacity that had always existed but never been accessible.
The Results
AI-optimised scheduling and constraint analysis transformed the operation:
13.5% improvement in throughput by relieving bottlenecks and synchronising production stages from milling to bottling.
56% reduction in idle time through AI-optimised scheduling that kept assets working productively.
80% elimination of labour excess hours with shift patterns designed around optimal production flow.
87% tank utilisation achieved while actually reducing the number of storage tanks required.
5% energy saving by eliminating inefficient production patterns and reducing idle equipment time.
Reduced acetation time through identified process improvements that had been hidden by scheduling complexity.
The manufacturer unlocked capacity they already owned—using AI to find the optimal schedule that human planners couldn’t see through the complexity of tank interdependencies and multi-stage production constraints.