Fiix by Rockwell Automation Upgrades Predictive Maintenance Software

Fiix by Rockwell Automation

Across North America, Rockwell Automation is well known throughout the industrial automation sector. Many factories use Rockwell Automation’s software to monitor and control their automated processes.

Fiix, located in Toronto, Canada, was acquired by Rockwell Automation in 2020 to supply a cloud-based computerized maintenance management system (CMMS). This software allows managers and maintenance staff to issue digital work orders, manage assets, inventory management, and analyze data. Recently, Fiix has added artificial intelligence to its CMMS software.

 

Risk assessments for individual assets can provide a breakdown of which machines require immediate attention. Image used courtesy of Fiix Software

 

Asset Risk Predictor from Fiix

Monitoring equipment and predicting breakdowns requires massive data collection, analytical tools, and complex algorithms. Even with this tall order of requirements, companies both large and small are requesting software to help prevent breakdowns or predict when a breakdown might occur.

The Asset Risk Predictor is an add-on to Fiix’s flagship CMMS software. The addition takes readings from sensors around the factory, stores the data, and provides configurable dashboards to the user.

Analytical tools are provided via the web interface, allowing maintenance staff to plan inspections based on hourly sensor readings or historical data. The dashboards show risk levels for each asset and can be categorized by recipe or operating environments. Custom alerts can be configured for specific users. This way, only staff who are in charge of specific assets will be alerted when the risk level is out of range. Asset risk level can also be predicted for each day, thus reducing inspecting or performing maintenance on equipment that doesn’t need maintenance.

 

Within a single asset, sensors can provide diagnostic and predictive data for daily analysis. Image used courtesy of Fiix Software

 

Teaching The AI Algorithm

AI tools have been around for some time, and we have seen the benefits of using them. A downside to AI tools is they need to be taught. Each AI analytical tool needs to understand the differences between a good and bad part. For predictive maintenance, the analytical tools will need to be supplied with references for what are considered ‘good’ or ‘bad’ characteristics.

If you are trying to predict when an electrical motor might fail, you would want to collect data on a new motor. As the motor wears, the analytical tools will need to understand how the data points to such a worn motor. This teaching process can take a long time; weeks or months, depending on the data collected. The Asset Risk Predictor claims to be able to learn about equipment failures in as little as seven days.

 

All equipment is crucial to a process, but risk analysis determines just how costly each individual breakdown might become. Image used courtesy of Unsplash

 

Factory Maintenance Strategies

Regardless of the size of your company, if you have invested time and money into automated equipment, you need to maintain that equipment regularly to ensure productivity.

A lack of clear failure prediction leads to creating some blanket maintenance schedule at whatever frequency the company can afford. This strategy causes unnecessary maintenance which can be expensive and still leads to unpredicted breakdown between service intervals.

Predictive maintenance software works by collecting data on equipment and learning about the system. When the system is running well, the data collected will reflect that. The AI algorithms analyze the data from each asset and try to predict when those assets might need maintenance. In the case of an electrical motor, vibration and current draw will both increase as the motor ages. The AI algorithm can trend those values and come up with a predicted time of replacement or service.

AI tools are certainly increasingly powerful and use advanced mathematics, but they are not foolproof. If the data collected is incomplete or inaccurate, the tools could give false predictions. The Asset Risk Predictor from Fiix and Rockwell Automation offers a fast learning time for AI tools, with the goal of saving you money within a few weeks instead of months.