How to Build a Better Predictive Analytics System

The predictive analytics system uses AI to determine potential machines breakdowns before the failure occurs.

Rob Spiegel

May 30, 2024

2 Min Read
OMNI edge predictive analytics
THK

At a Glance

  • Predictive analysis allows users to take bite-sized chunks of big data to determine the future behavior of machines.
  • The data can point out what users can learn from past events like machine breakdowns.
  • The goal is increased efficiency, reduced inventory management costs, and improved machine operating rates.

THK, Inc. developed the OMNI edge as a predictive analytics system. The system uses a secure communications network to analyze periodic real-time data of machine components and perform predictive failure detection. Predictive analysis allows users to take bite-sized chunks of big data to determine the future behavior of machines. The data can point out what users can learn from past events like machine breakdowns. It does this by using artificial intelligence to explore historical data, identify patterns, and answer the question, “What’s going to happen next?”

The technology of OMNI edge uses an algorithm and reference data to quantify the status of components, monitor their condition, and detect parts failures before they occur. This greatly increases optimal machine operation as well as uptime. The overall goal is to make maintenance more efficient, reduce inventory management costs, and improve machine operating rates.

We caught up with Atsushi Matsui, general manager at THK America, Inc., to get the details on how the predictive analytics system works.

How and why was the OMNI edge predictive analytics system developed?

Atsushi Matsui: As we enter the 5G and 6G era, further strides will be made in automation and robotization once Industry 4.0 environments are established. The speed at which a company can achieve predictive failure detection and preventative measures will be critical as machines become increasingly interconnected. In this situation, a key feature of OMNI edge is that, in addition to being installed on new machines, it can also be added to the machines customers are currently using and are more concerned about. Furthermore, gathering and analyzing large volumes of data will enable us to not only perfect the service, but also develop new products and provide upgrades.

What type of machine issues does the system identify?

Atsushi Matsui:

  1. OMNI edge for linear motion systems

  2. OMNI edge for rotary equipment

  3. OMNI edge for cutting tools

How does the system use AI to determine “What’s going to happen next?”

Atsushi Matsui:

  1. OMNI edge for linear motion systems

  2. OMNI edge for rotary equipment

i. Alert when abnormality is detected

ii. Predicted failure mode

iii. Maintenance recommendation

iv. Diagnostic
3. OMNI edge for cutting tools

Explain the extensive reference data that is used to quantify the status of components. How does it work and how was it developed?

Atsushi Matsui:

  1. OMNI edge for linear motion systems

  2. OMNI edge for rotary equipment

  3. OMNI edge for cutting tools

How does OMNI edge reduce inventory management costs?

Atsushi Matsui: Traditionally, the MRO parts inventory is either overstocked due to a fear of unexpected failure or understocked, which causes the production line to stop.  Either case results in unwanted, excessive costs.  OMNI edge provides users with an ability to predict failure of machine components.  This allows the users to effectively manage their MRO parts inventory by optimizing the inventory quantity and procurement timing. 

About the Author(s)

Rob Spiegel

Rob Spiegel serves as a senior editor for Design News. He started with Design News in 2002 as a freelancer and hired on full-time in 2011. He covers automation, manufacturing, 3D printing, robotics, AI, and more.

Prior to Design News, he worked as a senior editor for Electronic News and Ecommerce Business. He has contributed to a wide range of industrial technology publications, including Automation World, Supply Chain Management Review, and Logistics Management. He is the author of six books.

Before covering technology, Rob spent 10 years as publisher and owner of Chile Pepper Magazine, a national consumer food publication.

As well as writing for Design News, Rob also participates in IME shows, webinars, and ebooks.

Sign up for the Design News Daily newsletter.

You May Also Like