In 2020, my team at eyllo (https://eyllo.com) and I did a Proof of Concept to help a logistics company with inspection and maintenance of their locomotives. In order to provide step-by-step instructions, we deployed our assisted reality software that runs on smart-glasses to help technicians perform their work.

Context

The railroad company operates in Brazil, connecting the central part of the country with states in the south and southeast. Many disciplines are involved in servicing a locomotive. The main maintenance center is located in the south region. In this maintenance center, technicians are able to perform all necessary tasks to keep the locomotives operational. Technicians are also stationed in other locations along the routes and specialized technicians are dispatched from the main service center when necessary.

For this project, among the key performance indicators used to measure maintenance quality, we focused on two relevant ones:

• MTTR (Mean Time To Repair), that is, the mean time it takes to fix problems once the service center receives a notification of a failure.
• MKBF (Mean Kilometers Between Failure), that is, the mean distance a locomotive is able to travel, operating properly and generating revenue, before it fails and requires assistance.

Proof of Concept

For this project, we deployed some realwear smart-glasses with our proprietary software Vista in order to help with work flow execution of service orders and remote expert.

To reduce MTTR, in locations where connectivity was available, the field technician was able to make an audio/video call with a remote expert. The remote expert was able to see the situation of the locomotive, have a better understanding of the issue and assist the field technician.

In order to increase MKBF, we focused on reducing task variability. The use of step by step instructions to digitize the work flow allowed for a reduction of variability and improved quality.

Experiments

We designed and ran two experiments to assess the proper use of our solution to help reduce time to repair and task variability.

The first experiment involved using a remote expert located at the main service center to assist a technician in the field. The expert at the service center used a regular web browser on a personal computer. The field technician used our mobile application installed on the smart-glasses. The smart-glassed were connected via wi-fi to a cellphone working as an access point.

For the second experiment, we created step-by-step instructions to assist with the service orders. Together with the customer, we selected two workflows. We created the content with step-by-step instructions and uploaded the content to our solution. The technician would scan the QR code and follow the visual and audio instructions in order to properly execute the tasks.

Results

In this Proof of Concept, together with the customer, we were able to deploy an Assisted Reality solution in order to help with reducing task variability. We were also able to reduce the time it takes to repair a locomotive when there is an issue far from the main service center.