How Amazon Air Completed 2,914 Sessions of Machine Maintenance Training With Zero Downtime

Amazon Air trained maintenance engineers across 5 facilities, completing 2,914 sessions—all without any actual machine involved.

Amazon Air is the aircraft transportation wing of Amazon. It has a fleet of 100+ aircraft and operates around 250 daily flights that deliver packages across long-distances. Amazon Air is a critical part of the Amazon logistics network known for its fast and prompt deliveries.

Training for machine maintenance without using the actual machine

Integral to Amazon’s prompt delivery network are hundreds of conveyor belts across multiple Amazon Air warehouses. If a conveyor belt stops for a few hours, hundreds of deliveries become delayed.

Preventing downtime requires prompt maintenance and repairs from warehouse maintenance engineers. All of whom require quality training.

But training meant waiting for a conveyor belt to break down. Or practicing on an older, out-of-service conveyor belt.

Both limited the number of people Amazon Air could train and evaluate across a given period.

Integral to Amazon’s prompt delivery network are hundreds of conveyor belts across multiple Amazon Air warehouses. If a conveyor belt stops for a few hours, hundreds of deliveries become delayed.

Preventing downtime requires prompt maintenance and repairs from warehouse maintenance engineers. All of whom require quality training.

But training meant waiting for a conveyor belt to break down. Or practicing on an older, out-of-service conveyor belt.

Both limited the number of people Amazon Air could train and evaluate across a given period.

Simulating machine maintenance in virtual reality

Amazon Air partnered with AutoVRse to create a library of VR training for conveyor belt maintenance.

Based on Amazon Air’s existing documentation and input from their on-the-ground engineers, AutoVRe created 13 VR modules across two phases.

The 13 modules featured both training and evaluation, first teaching the engineers the correct procedures in a virtual environment and then testing their performance in the same environment.

Amazon Air partnered with AutoVRse to create a library of VR training for conveyor belt maintenance.

Based on Amazon Air’s existing documentation and input from their on-the-ground engineers, AutoVRe created 13 VR modules across two phases.

The 13 modules featured both training and evaluation, first teaching the engineers the correct procedures in a virtual environment and then testing their performance in the same environment.

Delivering measurable training at scale

The modules were built around specific training needs Amazon Air had:

Training that doesn’t cause machine downtime
The modules replicate real-world conveyor belt operations and maintenance procedures.

Different conveyor belt models
Phases 1 and 2 covered two different models of conveyor belts used in Amazon Air facilities.

Support training workers across facilities
A large number of workers across multiple facilities can be trained and evaluated within a short time.

Performance that’s backed by data
Every training session, in every facility, is tracked and recorded.

Achieving 2,914 completed training sessions

The 13 VR modules have become an integral part of Amazon Air’s training ecosystem:

2,914 sessions completed
5 facilities deployed in
99.4% pass rate in Phase 1
Supports onboarding and knowledge refresh

Meaning more engineers to perform proper maintenance. Issues get fixed faster. Downtime is kept at a minimum.

Scaling to more facilities and use cases

Prompted by the success, Amazon Air plans to scale the 13-module training library to cover 25 facilities.

AutoVRse and Amazon Air are in conversation about the next phase, which involves additional use cases and facilities.

Let’s talk about

your training

Talk to our team to learn how to implement VR training at scale