Improving Analytics Productivity Across the Enterprise

Improving Analytics Productivity Across the Enterprise
Improving Analytics Productivity Across the Enterprise
Plant managers are faced with the dual challenges of increasing production efficiency while maintaining aging equipment, and corporate executives must do the same across the entire manufacturing enterprise. In the process industries, subject matter experts (SMEs) are those end user employees with the expertise required to understand processes and equipment. They are typically engineers, but in some cases may be educated in some other technical discipline. Subject matter experts (SMEs) are critical to provide the expertise necessary for meeting these challenges, but they are in short supply.
 
Therefore, it is vital to increase SME productivity, which can be done through the effective application of digital technology innovations such as the industrial internet of things (IIoT), artificial intelligence (AI), and machine learning (ML). Further increases in SME productivity can be realized by supporting their efforts with business and domain knowledge supplied by vendor experts, such as those at Yokogawa, particularly when data is stored in the cloud for secure access by all relevant parties.
 
Production data obtained from the operational technology (OT) domain is the main, and often the only, source of data for many manufacturers. SMEs, operators, and other plant personnel use this data extensively to improve productivity, both at individual plants and across the enterprise throughout a fleet of operating equipment.
 
Increasingly, this OT data will be supplemented with facility and equipment data obtained via the application of information technology (IT) innovations, creating new opportunities to fuse OT production data with IT equipment data.
 
This will allow manufacturers and SMEs to progress from sensing to sensemaking by following these three steps.
 

From Sensing to Sensemaking

There are three steps for achieving digital transformation (DX). After first digitizing Information (digitization), sensing becomes sensemaking by digitizing data analysis (digitalization), and finally by realizing DX of corporate activities and organizational culture.
 
Digitization of information is simply digitizing visual or paper-based information. In this step, analog information is encoded and converted into digital information so it can be easily stored and processed. This is most often done by automating manual data acquisition activities using data collected from sensors connected to wired and wireless networks, which can be referred to as sensing.
 
Sensing is already common for the production data required to automatically operate plants in real-time. But equipment data, which includes all of the other data which could potentially be used to improve production but is not required for real-time automated operation, is often not sensed and is instead collected manually. This manual data collection presents a number of problems including high costs, exposure of workers to dangerous or hazardous environments, infrequent readings, and errors introduced from entering readings into computer-based platforms. Due to these and other issues, equipment data is lacking for most manufacturers, hindering digitization efforts.
 
Sensemaking in the Digitalization step gives meaning to the data acquired by digitization. The end goal for manufacturers isn’t collecting data, but to instead create value from this data. For example, by monitoring data trends, new metrics can be generated for indicating signs of abnormality. This type of data analysis can be performed automatically by using AI and ML, freeing SMEs from some of the tasks associated with sensemaking.
 
DX is the final step, a transformation of corporate activities and organizational culture to bring about a change in the way companies, businesses, and people interact to improve the operational performance of plants and fleets. In the DX step, further value creation can be made by organically integrating OT and IT functions and data. For example, equipment data concerning the condition and performance of equipment can be digitized and fed to the production process for improved optimization.
 
Fortunately, new technologies are facilitating the sensing to sensemaking process.
 

Improving Digitization

The digitization step requires improved sensing to:
  • Measure various physical quantities accurately to ascertain equipment condition with sufficient data quality and fidelity
  • Digitize equipment health and performance data
  • Simplify installation and operation of sensors
  • Provide long-term operation of sensors in often harsh environments with minimal required maintenance
 
Yokogawa developed the low-cost wireless Sushi Sensor, a system for connecting to industry standard low-power wide-area networks to address these and other issues. Sushi Sensor uses the LoRaWAN open communication standard, which is supported and promoted by the LoRa Alliance and its 500 plus IoT-related companies and end user members. Even in a typical process plant environment with various physical obstacles, long distance communication with a radius of 1 km is possible, without the need for repeaters and wireless routes (Figure 1).
 
  
 
Figure 1: Sushi Sensor features make them ideal for deployment in typical plant environments.

These sensors are battery-powered, so no wired infrastructure of any kind is required. Installation is simple, for example the vibration sensor is installed by either a fixing screw or magnet. With data updates of once per hour, intervals between battery replacement exceed four years in most applications. These sensors are rated IP 66/67 dustproof and waterproof, and they are also explosion-proof.
 
These types of sensors and the LoRaWAN network are very different from the traditional instrument networks, such as ISA100 and WirelessHART, deployed in many plants for collecting OT production data.
 
The LoRaWAN sensors are smaller and lower in cost than those used in traditional instrument networks, have very low power usage, don’t required much in the way of network bandwidth, and can communicate over much longer distances.
 
When these types of sensors are used in facility and equipment monitoring applications to replace manual monitoring, SME productivity is improved because data is collected more frequently with higher fidelity and fewer errors. More and better data also improves upon the next step, digitalization.
 

Better Digitalization

The digitalization step is where sensemaking is performed, and it requires:
  • Ability to do advanced data analysis
  • Monitoring trends for effective utilization of sensed data for equipment maintenance
  • Functions to detect signs of abnormality before equipment fails
  • Fault diagnosis
 

Figure 2. Operator rounds can be replaced with online monitoring using Sushi Sensor.
 
Using this data, the condition of each item of rotating equipment was monitored. Results were used to schedule maintenance only when needed, resulting in fewer unplanned outages, longer time intervals between repairs, little or no reactive repairs, and reduced costs.
 

Detection of Ball Bearing Abnormality Using AI and ML

Sushi Sensor was installed on an item of rotating equipment to measure vibration and transmit this data over LoRaWAN. As shown in Figure 3, a scored label, a type of health index indicator, was automatically created using an AI/ML algorithm showed there was a sign of abnormality three months before failure. By combining the sensor data with and the AI/ML algorithm results, it was possible to determine signs of the ball bearing abnormality, a predictive factor for rotating machine failure.
 

 
Figure 3: Advanced predictive maintenance can be performed using Sushi Sensor and AI.
 
This is an example of automated sensemaking in which data digitized by sensing was able to create new value using AI/ML. SME productivity was increased by freeing them from the task of manually analyzing data to predict rotating equipment failures.
 

Conclusion

SMEs are critical to analytics efforts, but they are in short supply at most manufacturing companies. Intelligent application of digital technologies can increase SME productivity significantly, both for a single plant and across an enterprise. When data is stored in the cloud, the business and domain knowledge of vendor experts can be applied to supplement end user SME efforts, further increasing their productivity and effectiveness.
 
Figures all courtesy of Yokogawa

About The Author


Takayuki Sugizaki is a manager at Yokogawa’s CX Strategy Department of Information Technology Center. He joined Yokogawa in 2003 and is responsible for promoting wireless products and solutions worldwide.
 

Click Here for More Information

Did you enjoy this great article?

Check out our free e-newsletters to read more great articles..

Subscribe