The solution
Jugos S.A. operated a large concentrated apple and pear juice plant where weekly schedules had to reconcile incoming fruit forecasts, commercial commitments, tanks, cleaning days and product quality. A simple calendar plan could say what to make, but not whether the semiprocessed pools had the right sugar and acid profile.
The team built two complementary mixed-integer decision-support modules. PLANNER batched orders, scheduled production and tracked semielaborated juice pools; SIMOPT optimized plant operating conditions and computed specific costs so the planner could compare scenarios economically.
The neat move was to treat scheduling and process economics as one feedback loop. The tool produced feasible schedules, target conditions and long-term guidelines, while letting managers test scenarios before the fruit and tanks locked them in.
What it achieved
Plan the batch and the blend chemistry together
Why it worked
It included cleaning, rest days, raw-material availability and storage constraints in the same planning layer.
Pooling logic made hidden quality constraints, such as sugar and acid content, part of the schedule.
The simulator supplied cost information rather than leaving planners to guess which feasible schedule was best.
What can be applied
When production depends on both capacity and product chemistry, the clever schedule is the one that keeps the intermediate pools visible.
Aftermath
The 2020 INFORMS Journal on Applied Analytics article documented the deployed tools as decision support for production guidelines and scenario studies at Jugos S.A.'s concentrated fruit juice plant in Rio Negro, Argentina.
FOLLOW THE EVIDENCE