Predictronics had a fruitful experience at The Reliability Conference 2019 in Seattle, Washington.

The Reliability Conference is an annual event that brings together industry professionals, giving them the opportunity to share their expertise and the latest innovations in reliability and asset management.

Senior Data Scientist Aaron Shelly hosted a workshop, A Combined Analytics and Domain Knowledge Approach for Predictive Monitoring of Control Valves, on the second day of the conference. In his session, Mr. Shelly discussed the steps for creating an effective and impactful predictive solution.

Featured at the conference was our unique end-to-end predictive analytics and industrial AI solution, PDX, which collects, analyzes, monitors, and visualizes data to pinpoint potential failures and allow for planned downtime and maintenance. Mr. Shelly presented one of our PDX case studies which demonstrates the real-world application of predictive monitoring algorithms to detect early faults in control valves.

As valves are an integral component to various industrial applications, including water, air, and gas systems, as well as food and beverage production and biopharmaceutical processing, the impact of this solution is both broad and substantial. Maintaining a competitive edge is key, and a significant opportunity exists for OEMs to offer predictive services to end-users, as well as to adopt similar technologies in their own production facilities.

The approach of our case study correlates the health index with common valve issues, such as travel deviation and increased friction, forecasts the degradation and failure of unhealthy valves, and warns the user when health thresholds are crossed, preventing unnecessary downtime and reducing cost and wasted resources.

There are numerous other benefits to utilizing an automated data analysis method over human inspection. The traditional manual evaluation method for analyzing valve performance requires a great deal of human manpower and resources, along with vast expertise and domain knowledge.

Predictive solutions, such as product quality and health monitoring, affect both the manufacturers of valves, as well as end-users. From a reliability standpoint, product quality is the key to manufacturing, as that will affect the reliability of products in the field. For end-users, predictive systems are the key to reliability, informing reliability practices and maintenance schedules.

There were numerous other noteworthy speakers that attended the event, discussing important topics and challenges within the field of maintenance and reliability, including representatives from Tesla, Intel, National Instruments, Saudi Aramco, the U.S. Army Corp of Engineers, Tacoma Power and more.

Below are Mr. Shelly's top four takeaways from the conference:

1. You Can't Talk About Reliability Without Talking About Asset Condition Based Maintenance

One highlighted theme at the conference was establishing a better understanding of the root cause of component failures within an asset. Root cause determination is the key to accurate and timely decision-making, not only in maintenance, but also in process design. Identifying the root cause requires knowledge of the possible failure modes and the parameters exhibited during those failures. A failure mode consists of the location of the problem, i.e. the component, a description of the problem, and the cause of the problem. Recording this information through failure codes and maintenance management systems does not necessarily provide all three types of information. The proper labeling of failure mode information is important for performing subsequent criticality analysis, deciding which sensors or parameters would be the most suitable for predictive monitoring solutions, and determining the root cause of the failure. Ultimately, implementing countermeasures that reduce the risk of the failure and/or prevent it from occurring altogether would be the most ideal solution.

2. Predictive Analytics and Machine Health Monitoring Are Just One Part of the Solution

Being able to monitor the health of machines and predict when they will fail is essential. Reliability, however, helps to stop machines from failing, period. This is accomplished through additional training and better policies, including improved lubrication practices, periodic machine shutdowns, maintenance, and more. Data analysis gives manufacturers and solutions providers the evidence they need to make better maintenance and reliability decisions, creating a feedback loop that helps determine why a machine would fail and how it can be prevented.

3. Healthier Machines Lead to A Safer and More Productive Work Environment

Reliability is one of the main contributors to a safe working environment. Many attendees at the conference expressed how poor machine reliability and recurring failures can actually lead to maintenance accidents and deaths within manufacturing facilities. For many, learning about reliability maintenance isn’t about making more money, reliability means the difference between life and death.

4. It’s Important to Have Everyone Involved When Developing New Reliability Solutions

Another topic of importance at the conference was leadership in reliability. Achieving reliability is only possible when there is a culture of reliability, with everyone participating to make the production process and workplace efficient and safe. Reliability is not just a maintenance issue; it should be a company pillar with an emphasis on collaboration amongst all employees, from the highest decision-makers to the machine operators on the floor. If business executives cannot be shown the value of innovative solutions, there will not be a company-wide push to establish new initiatives for change that would better not only production, but also the work environment.