I Am So Proud Of This! Predictive Maintenance

Andreas Kretz Blog 1 Comment

Ever wish you could get an easy answer to what is Big Data? Get the Big Data Checklist, totally free, along with weekly blog tips delivered directly to your inbox.

Just collecting data is not enough! 

Today I am going to show an awesome use case that combines Big Data and Data Science: Bosch Rexroth’s ODiN platform (Online Diagnostic Network)

It helps customers look into the future and detect machine failures before they even happen.

How is this possible? With predictive maintenance.

Find out below what’s my role in that and why I am so proud of this.

What is Predictive Maintenance?

Until now machine maintenance is usually done by a fixed schedule. If something goes wrong before another maintenance is scheduled your machine breaks down.

You know this from your car’s maintenance intervals. Mine currently says 17 thousand kilometres or 10 months.

If you have a problem in between you either break down, or you limp back to the repair shop to have the problem fixed.

Breakdowns often have the consequence that greater damage happens to the machine. In such cases repairs take time, cost a lot of money and the machine is not able to produce.

The goal of predictive maintenance is to predict breakdowns before they happen. It basically tells you when you need do maintenance to keep your machine at optimal performance.

How do you do that?

With a lot of Data and clever analytics.

How is Rexroth doing predictive maintenance?

Rexroth is doing predctive maintenance in six steps:

Step 1: Data acquisition

To do predictive maintenance you need data. One issue is that a lot of older machines do not produce any data or not the right data.

To tackle this Rexroth is offering and installing sensor packages and data acquisition units to acquire data.

Step 2: Storage

The acquired data gets send through a secure connection to the ODiN cloud system. There the information gets stored and processed.

Step 3: Predictive Analytics

Once the data is ingested by ODiN, the machine learning kicks in. After a teach-in period odin creates health predictions for the machine and it’s components.

Step 4 and 5: Maintenance Recommendation & Measures

Through this clever analytics ODiN creates maintenance recommendations of when to service the system. Experts then define measures and give advice to customers how to execute them.

Step 6: Intelligent drives and controls

Installing a lot of sensors is not optimal. It needs time and money to make the machine IoT ready.

Rexroth is simplifying how to enter the predictive maintenance world. By offering drives and controls that have predictive maintenance built in.

Check out this shot of Rexroth’s customer magazine drive & control to learn more:

What is my role in in this venture?

As a Big Data Solution Architect I am responsible for the ODiN cloud platform. From conception to implementation and production.

I am responsible for Big Data storage solutions as well as stream and batch processing pipelines. I act as the link between the data scientists needs and the big data engineering team.

As head of user experience and user interface development I am making sure that it is awesome to work with ODiN.

I am also responsible for API design and development and platform security.

Why I am so proud of this

I am always preaching that big data and data science are not only hype words. If you have a vision and combine big data with data science you can create immense value for customers.

This is exactly what we have done here at Rexroth.

How much value? Check out page 18 in the drive & control magazine. KREIBURG GmbH & Co. KG who runs a rubber plant writes about their experience with Rexroth predictive maintenance.

You can find all the above material and more infos about Bosch Rexroth in the drive and control magazine: Link

I am very happy and grateful to be a part of this 🙂

Do you know work on something as awesome as this? Please share it with me and the community in the comments section below.

If you like this article please share it with your friends on social media.
 

Comments 1

  1. Andreas I would love to explore the application for this for Utility Customers (appliance level monitoring for energy usage and state of health)and then Utility as a customer.

    The challenge is in the first step you identified i.e data acquisition.

    Happy to chat and explore.

Leave a Reply

Your email address will not be published. Required fields are marked *