Case study · Coca-Cola FEMSA · Operations, Mexico

The simulator that taught order picking without risking a single delivery.

Coca-Cola FEMSA needed operators to master order picking before touching a real order. WASD Studio built a web 3D simulator that replicates actual purchase orders from their ERP, with adaptive difficulty. The program recorded 44% fewer picking errors and a third less training time, and it is now the flagship piece of a program of 5+ simulators.

Reply within one business day · No commitment

TRUSTED BY THE MOST DEMANDING ENTERPRISES IN MEXICO

Coca-Cola FEMSA
Alsea
Philip Morris International
Teleperformance
Nissan
BASF
Unity
L'Oréal
Grupo Splitel
IT Pharma

A picking mistake doesn't stay in the warehouse. It reaches the customer.

In a bottling and distribution operation, bad picking gets paid for three times: in the loading block, in the delivery that doesn't go out and in the sale that falls through. Coca-Cola FEMSA onboards around 2,000 operations employees a month, so the critical process has to be taught over and over, without slowing the operation and without using real orders as practice material.

  • Practicing on real orders means risking loading blocks that hit sales and delivery times.
  • Turnover forces the same process to be retaught continuously, not once a year.
  • Floor coaching depends on an instructor being free and on every shift explaining it the same way.

What we built

A simulator wired to the orders the operation already receives

The flagship piece of the program is a web 3D Picking simulator that replicates actual purchase orders from Coca-Cola FEMSA's ERP. Operators don't practice against an invented catalog: they practice against the kind of order they'll get on Monday, with the same product combinations and the same traps. That fidelity is what makes the practice transfer to the floor instead of staying in the classroom.

Adaptive difficulty: the simulator raises the bar as people master it

The simulator adjusts difficulty to each operator's performance, so whoever already has the basics moves on and whoever struggles repeats the scenario that beat them. It's adaptive learning applied to an operational process rather than a linear course everyone finishes at the same pace, and it's why the same content works for people with very different experience.

Browser, not headsets — and that was a design decision

It was built to run in the browser the operation already has, not on virtual reality headsets. When one critical process has to reach tens of thousands of people across many sites, hardware friction is the real bottleneck: every headset you have to buy, charge, sanitize and administer is one more person who didn't practice. Recommending the cheaper format when it fits is part of the job.

From one piece to a program of 5+ simulators

Picking opened the door and the program grew past five simulators now covering manufacturing, commercial and pre-sales, and HR: Delivery and Warehouse, Area Virtualization, Leader-Ship and the HR Diploma. That's the evidence we care about most in this case: it wasn't a pilot that impressed and got shelved, it was a program that kept being extended.

Measured results

What the operation recorded

44%fewer picking errorsIn the put-away process, after training on the simulator.
1/3less training timeReduction in the time needed to get one person ready.
Tens of thousandsof employees trainedCumulative across the program, without stopping operations.
5+simulators in productionManufacturing, commercial and pre-sales, and HR.

These are operational and learning indicators from the program, not a promise of financial return: WASD Studio does not publish business ROI percentages.

How it was built

  1. 01

    Capture of the real process

    We mapped put-away exactly as it happens on the floor, with the people who perform it, so the simulator wouldn't contradict the operation.

  2. 02

    Wiring to the ERP orders

    We replicated actual purchase orders from the system, which is what separates a generic exercise from practice that transfers.

  3. 03

    Design of the adaptive difficulty

    We defined with the client how the challenge scales with performance, and which scenario repeats when someone fails.

  4. 04

    Rollout and growth of the program

    We launched in the browser to remove hardware friction, and the following simulators were built on that base.

WHAT OUR CLIENTS SAY

What we value most about WASD Studio is their capacity for technological innovation. They've made training management with our internal teams enormously easier — something that used to be quite complex to coordinate.

Alexzandra GranadosTraining & Learning ManagerCoca-Cola FEMSA
Coca-Cola FEMSA

FREQUENTLY ASKED QUESTIONS

Why a web simulator instead of virtual reality?

Coca-Cola FEMSA's Picking simulator runs in a browser because the goal was volume, not immersion. When one critical process has to reach tens of thousands of people across many sites, hardware friction decides how many people actually practice. WASD Studio builds in virtual reality when practice needs physical presence — machinery that can't be stopped, safety procedures — and in the browser when the constraint is scale.

What does it mean that it replicates real ERP orders?

The simulator is fed the kind of purchase orders the operation receives in its own system, instead of invented exercises. Operators practice against the same product combinations and the same traps they'll meet on shift, which is what makes the practice transfer to the floor rather than stay in the classroom.

How is each person's performance measured?

Every attempt inside the simulator is recorded, so the training lead sees what each operator did and where they failed, not an attendance sheet. That reading is what makes it possible to decide who can already take real orders and who needs to repeat a scenario.

How long does a simulator like this take to build?

A typical onboarding or operational training project takes 10 to 14 weeks, depending on how many processes are simulated and how much fidelity the environments need. Coca-Cola FEMSA's program wasn't built in one go: it started with Picking and grew past five simulators as the operation asked for new processes.

Does this apply if my operation isn't beverages?

The mechanic works for any critical process that is learned by doing and expensive to practice for real: put-away, logistics, equipment start-up, floor service. What gets adapted is the process and the system the data comes from; what stays is the logic of practicing against real cases with difficulty that scales.

Tell us which critical process is costing you errors, and we'll come back with a quote within one business day.

Tell us your training challenge and we’ll get back to you within one business day.

16+ Years50+ CompaniesUnity Award Winner

Reply within one business day · No commitment