A renowned Brazilian appliance company, specialized in manufacturing small appliances, air treatment products, and kitchen items, faced a classic challenge of modern manufacturing: how to distribute hundreds of products across multiple production lines while respecting capacities, compatibilities, and demand targets, all with speed and predictability. The process, once manual and run on spreadsheets, took days of work and relied heavily on each planner’s individual experience. Partnering with BlueMetrics, the company implemented a structured production optimization system that combines mathematical algorithms, scalable cloud architecture, and an interactive analytics dashboard.
Overview
In large-scale industrial operations, production allocation decisions have a direct impact on costs, deadlines, and service levels. That was exactly the challenge faced by one of Brazil’s leading appliance manufacturers, with a broad portfolio that includes small appliances, air coolers, and kitchen items.
The company ran multiple production lines, each with its own characteristics, such as:
- Different production capacities
- Variable speed by product (PPH, pieces per hour)
- Compatibility constraints between products and lines
In this context, monthly planning had to answer critical questions, such as:
- Which product should be made on which line?
- How much of each item to produce to meet demand?
- How to maximize the use of available capacity?
- How to reduce unnecessary changeovers and setups?
With hundreds of SKUs and high demand variability, small inefficiencies quickly added up, creating meaningful impact on the operation.
Problem: operational complexity and spreadsheet-based decisions
Before the solution went live, production planning was done entirely in spreadsheets. Beyond being manual, the process relied heavily on each planner’s individual experience.
As the portfolio grew and operations became more complex, the model began to show serious limitations. The main challenges included:
- Full days spent building the monthly plan
- Manual analysis of hundreds of product-by-line combinations
- Imbalance across lines (some overloaded, others idle)
- Difficulty simulating alternative scenarios
- Frequent rework when demand changed
- Lack of structured traceability for decisions
The impact went beyond the operational level and hit the business directly:
- Loss of production efficiency
- Added costs from overtime
- Delays caused by constant replanning
- High team effort to keep the plan stable
It became clear that the problem was not just about speeding up the process, but about building an approach that could mathematically optimize decisions, bringing consistency and predictability.
Solution: intelligent optimization with scalable architecture and data-driven decision-making
To solve this challenge, BlueMetrics built a complete production optimization system that combines advanced mathematical modeling, modern cloud architecture, and an intuitive analytics interface.
Optimization with two complementary approaches
The solution was built on two strategies that work together:
**1. Linear programming (mathematical optimization model)
This approach seeks the global optimal solution, considering multiple variables and constraints at once. Its main capabilities include:
- Maximizing demand fulfillment
- Respecting each line’s production capacity
- Accounting for the specific PPH per product and line
- Ensuring operational compatibility
This layer is ideal for deeper analysis and strategic planning.
**2. Heuristic (greedy) algorithm
The heuristic model was designed for speed and flexibility, allowing near-instant simulations. It works from rules such as:
- Prioritizing products with less allocation flexibility
- Directing production to the most efficient lines
- Quickly generating planning alternatives
Combining these two approaches struck the right balance between mathematical precision and operational speed, supporting more robust decisions without slowing down response time.
Interactive dashboard for decision-making
To make the process accessible and actionable, the team built an interactive web environment that turned planning into a dynamic decision-support system.
In this environment, users gained access to:
- Detailed views of production allocations
- Real-time performance indicators
- Overall demand fulfillment rate
- Capacity utilization by line
- Comparison across different algorithms
- Export of executive reports
With that, planning shifted from static to exploratory, allowing quick adjustments and better-informed decisions.
Modern, scalable architecture
The solution was built on a robust cloud architecture, ensuring performance, security, and room to grow.
The main components include:
- Serverless API for on-demand processing
- Managed SQL Server database
- Secure authentication with access control
- Infrastructure as code for standardization
- Structured logs and full auditing
This technology base ensures:
- High availability
- Scalability as the operation grows
- Enterprise-grade security
- Easy maintenance and evolution
Results
The rollout drove meaningful impact across several dimensions of the operation.
Productivity
Planning that once took days is now done in minutes. On top of that, the company can now:
- Replan quickly when demand changes
- Sharply reduce manual rework
Production efficiency
Optimization brought much smarter use of available resources:
- Better balance across lines
- Greater use of installed capacity
- Less production fragmentation
Decision quality
Decisions are now based on objective, measurable criteria:
- Planning driven by mathematical models
- Full traceability of allocations
- Continuous monitoring through structured metrics
Scalability and analytics maturity
Beyond the immediate gains, the company advanced its analytics capability:
- Infrastructure ready to expand
- A solid foundation for future machine learning initiatives
- A stronger data-driven culture
Conclusion
This project shows how the structured application of mathematical optimization, paired with a modern architecture, can turn production planning into a genuine strategic lever.
By replacing spreadsheets and gut-feel decisions with an automated, auditable, data-driven system, the company gained not only efficiency but also predictability and control.
More than speeding up calculations, the solution brought intelligence to the production process, turning operational complexity into sustainable competitive advantage.