🔍 Article co-created with Francisco Estrada, General Manager at RM Sheet Metal México (🇲🇽 Mexico), and Augusto Unzaga, host of Keru Podcast.
🎙️ This article was inspired by Episode 31 of Keru Podcast.
Keru Podcast, created by Globalita.io, is a series of conversations (in Spanish) with industrial leaders from Latin America about management, technology and decision-making in complex environments.
How many more SKUs can your plant handle?
Managing an industrial operation with 20 products can be challenging. With 200, it gets tense. But when the portfolio exceeds 1000 active part numbers, planning, shop floor control and cross-area coordination begin to collapse… if there is no method.
This is the case of RM Sheet Metal México Francisco Estrada, where production spans multiple processes, families and technical variations. A mix that creates business opportunities, but also operational complexity.
“You can’t use the same control system for five products as for a thousand. If you don’t evolve the system, the system drowns you.”
High-mix industrial production poses a strategic dilemma: how do you stay efficient when demand, model, sequence or volume change every day?
High-mix industrial production: more variety, more risk
In multi-product and multi-family contexts, the typical problems multiply:
- Schedules redone daily because of customer changes.
- Unnecessary work-in-process inventory that blocks flow.
- Machine adjustments that steal time and focus.
- Small orders that require the same effort as large ones.
- Overloaded staff jumping between tasks without stabilizing any of them.
This unstable environment generates scrap spikes, missed deliveries and team burnout. But the cause is not variety itself, but the lack of a system to manage it.
“When everything changes all the time, your system has to be the thing that doesn’t change.”
Multi-family portfolio management: stop improvising
Francisco Estrada sums it up clearly:
“If you keep planning with Excel when you have more than a thousand part numbers, you’re planning badly.”
Multi-family portfolio management requires tools for grouping, standardization and prioritization. It’s not about doing everything, but about doing it with industrial logic.
This means:
- Classifying by production families, not just by customer.
- Defining fixed windows for similar products (leveling).
- Grouping set-up changes with SMED logic.
- Training with One Point Lessons (OPL) to accelerate multi-skilling.
This approach doesn’t just organize the flow. It also allows the team to anticipate, understand and execute without depending on a boss shouting priorities every hour.
Tools for maintaining efficiency in high-mix industrial production
When a plant faces hundreds of products, dozens of routings and small, changing orders, operational efficiency in complex environments no longer depends on effort, but on system design.
RM Sheet Metal México Francisco Estrada points this out clearly:
“The problem isn’t having a thousand part numbers, the problem is treating them all as if they were the same.”
That’s why, in this type of factory, it is essential to apply specific methodologies to level, stabilize and standardize the operation without holding back customization.
1. Heijunka: leveling so you don’t lose
The Japanese concept of Heijunka proposes leveling production not according to total demand, but according to a stable sequence that maintains the operating rhythm.
In high-mix plants, this makes it possible to:
- Avoid load peaks in certain processes.
- Distribute product variety in small, regular batches.
- Design production cells that can absorb flexibility without disorder.
The Heijunka Box is a visual tool for scheduling these sequences logically, allowing the operator to focus on executing, not on guessing priorities.
“It’s not about producing faster. It’s about producing without friction,” notes Francisco Estrada RM Sheet Metal México.
2. Advanced SMED: changing over without losing efficiency
Single-Minute Exchange of Die (SMED) is a methodology for drastically reducing tool changeover or adjustment time between products. But in high-variety environments, traditional SMED isn’t enough. An advanced approach is required, including:
- Smart grouping of models with similar adjustments.
- Redesign of tooling to minimize manual intervention.
- External presetting of variables before the changeover.
- Visual standardization with quick guides.
Well-executed SMED makes it possible to switch from one product to another in minutes, without losing quality or time. At RM Sheet Metal, this logic is vital to keep the delivery promise amid extreme variety.
3. Grouping matrices: bringing order to chaos
When variety is excessive, a grouping matrix can become your best ally. It involves classifying part numbers according to:
- Required processes.
- Necessary adjustments.
- Demand frequency.
- Batch size.
- Critical or standard customer.
This matrix makes it possible to define:
- Logical production sequences.
- Order prioritization.
- Common routings for training operators.
- Shared control points.
What looks like “infinite variety” is often just a lack of strategic grouping.
4. One Point Lessons (OPL): teaching exactly what’s needed
In an environment where the operator must adapt to many parts, machines or customers, traditional training doesn’t scale. That’s where the OPL (One Point Lesson) methodology comes into play.
An OPL is a quick, visual lesson that:
- Addresses a single point of knowledge.
- Uses real photos or diagrams from the plant.
- Is taught peer-to-peer or by line leaders.
- Focuses on critical tasks for the changeover.
Applying OPL makes it possible to maintain multi-skilling without overloading the operator. It also builds a knowledge bank that grows with each new model.
“Operational knowledge can’t live only in one person’s head. It has to be in the system,” says Francisco Estrada.
5. What you can apply in your plant tomorrow
🔹 Design a grouping matrix
Review your active part numbers and classify them by process family, set-up type or demand.
🔹 Implement a Heijunka pilot
Choose a critical line and level its sequence into small, stable batches. Observe the impact on flow and operational load.
🔹 Redesign your changeover checklist with SMED logic
Separate what can be done before, during and after the adjustment. Measure how much time you gain.
🔹 Create your first series of OPLs
Identify the most common errors and create a visual lesson with photos showing how to do it right.
🔹 Train your shift leaders in high-mix logic
Not all products carry the same weight. Teach them to prioritize based on value, not just urgency.
Final reflection
Multi-family portfolio management is not a curse. It’s an opportunity… if it’s handled with method.
Many factories fail not because they produce a lot, but because they didn’t design a system capable of absorbing that variety with order, visibility and continuous learning.
RM Sheet Metal México Francisco Estrada shows that the solution is not to reduce the catalog, but to learn to manage it with industrial logic.
“No product is too small if it enters the system with method. What kills efficiency isn’t variety. It’s chaos.”
What about your plant? Is it ready to grow in variety without losing stability?
🔁 Keru is a podcast (in Spanish) created by Globalita.io, where we talk with the people who lead Latin American industry from the inside.
Francisco Estrada is General Manager at RM Sheet Metal México, with more than 10 years of experience in planning, production control and end-to-end supply chain management. His track record includes leadership roles at companies such as Safran, Jabil and RIPIPSA, where he has coordinated engineering, purchasing, quality and planning teams in high-mix manufacturing environments. A specialist in demand synchronization, key indicators (OTD, scrap, schedule attainment) and multi-family portfolio optimization, he combines technical skills with strategic vision to maintain efficiency without sacrificing operational flexibility.
View profile and episode →
