Classification of ROM (Run-of-Mine) based on the characteristics of the ore (such as lithology) considering the impact on process performance and the quality of the final products (such as granulometry and concentrations).
Optimization of the use of the ROM classified in an integrated way with the processes and delivery planning.
Optimization of process variables (separation, washing, concentration) appropriate to the type of ROM used and the final product. The goal is to guarantee production with the necessary quality in the right quantity and at the right time.
Optimization of material flow considering the various production routes, maximizing productivity and reducing production and transportation costs, including inventory movement.
Planning of maintenance downtimes to reduce the impact on delivery times and ensure compliance with the contract specifications.
Inventory management optimized by quality, within its capacity limits, which aims at meeting demands and adapting to process fluctuations. The inventory level is the main variable that used to anticipate maintenance at a production facility or a peak in demand.
Blending of the various products and loading optimization that maximizes the use of the qualities available in the short and medium-term.
Compliance with quality specifications reducing sales penalties.
Selection of combined clients/contracts that maximizes profit. The model considers all possible decision levers to select the most profitable sales portfolio.
Track process indicators and stock levels in the short and medium term.
Centralize the integrated vision of planning in a single collaborative tool.
Compare the optimized results in different scenarios.
Import data from other systems in an automated and secure way.
This solution is implemented during a project conducted by Cassotis consultants. The work combines Operational Research techniques with Machine Learning to represent the production process and optimize decision making. Using clustering and classification techniques, Cassotis develops predictive models of available raw materials and their interaction with process variables. These relationships found in the data feed the mathematical model. The final solution is responsible for finding the optimal configuration of the decision variables to maximize your profits.