As “lower forever” becomes a potential economic norm for the process industries, executives are taking a hard look at asset maintenance, and more holistically at Asset Performance Management, for efficiency and bottom line improvements.

 

Historically a cost center, asset maintenance has been left mainly to preventive work and managed by maintenance departments. That view is quickly fading as these areas are undergoing a technology revolution. Advances across mobile, sensors, big data analytics and machine learning have matured in both scalability and uptime for proactive maintenance and asset management strategies to become reality.

This maps to larger trends across the industrial landscape (see McKinsey graphic below), where we’re seeing large advances due to similar technology convergences, from driverless cars to advanced robotics in manufacturing. Indeed, we are on the cusp of a new industrial revolution, and it’s become clear that bringing these technologies — with data science at its core — into the mix is vital to future process industry productivity and growth.

 

For the process industries, we see this new industrial revolution taking place largely through what we call the “science of maintenance.” As might be expected, it relies heavily on big data analytics and data science. The science of maintenance drives Asset Performance Management by leveraging historical and real-time operational data, processed via algorithms, to model failures and head issues off before they happen.

 

The recommendations from Asset Performance Management systems determine both what action is needed and when, at a system and individual asset level, to maximize uptime and performance. A plant doesn’t have to risk losing a year’s worth of value over a few days in an unexpected shutdown, or suffer failures due to operational changes that overly stress systems.

As executives embrace this approach to asset management and maintenance, what does it mean for the current and future generation of engineers that are keeping the systems running? Are they ready to embrace these technology changes, or is there a skill mismatch that will limit or derail investment in these new approaches to maintenance and asset management?

 

There’s no doubt these strategies, which leverage massive data sets, require data science and engineering talent. That talent is scarce, and even the largest IT vendors are scouring for qualified hires. As stated by ARC Advisory Group, process industry organizations are loath to add headcount and aren’t “looking to hire a bunch of data scientists.” This is logical, as organizations want these technologies to automate tasks — and not inflate staff budgets and associated overhead costs.

                                                   

The good news is there are trends on both the employee and provider side that can take advantage of data science as applied to maintenance and asset management.

                                                                                                                                     

The path to applying data science to the process industries actually dates back to the late 1970s and early 1980s, when organizations started to upgrade their plant control systems from analog to digital, enabling the capture of second and minute data from thousands of sensors across these plants. When stored, this data enabled the development of higher-level applications such as advanced process control and optimization to create value through real-time computation to drive these plants closer to their operating limit. Immense value has been created over the last 30 to 35 years in the process industries from advanced solutions. 

An even bigger opportunity is now available by leveraging that same continuously stored data to derive insights and apply it in a predictive and prescriptive manner to improve the reliability of process and mechanical equipment. These predictive and prescriptive capabilities are enabled today by the computational and storage capacity available in the cloud through high-performance computing and data lakes.

 

In addition, a new generation of engineers are pushing the boundaries of what’s possible due to the ease of use and capabilities provided by the software, as well as computing capacity.

Many of the leading consumer and back-office applications and services in daily use today incorporate AI, the cloud and large data sets. The process industry user base is ready to migrate their habits to a more proactive and productive approach. 

 

This places the onus on software and service providers like AspenTech to appropriately integrate data science capabilities into our products. We’ve made significant progress on that front, including a variety of data scientists hired through acquisitions such as Mtell. We’ve seamlessly incorporated these capabilities into a platform that brings data science and analytics value across all phases of design, operations and maintenance. This approach takes the burden off customers and empowers engineering staff to take advantage of data science’s benefits.

Ensuring you have access to data science capabilities either internally or through a vendor — or, best case, from both — is a business imperative. It will position the organization at the forefront of the next industrial revolution, which is set to accelerate like no other phase of growth we’ve seen to date.

Learn more about the trends driving the future of maintenance in my recent executive brief, “The Time is Right for Optimum Reliability: Capital-Intensive Industries and Asset Performance Management.”