By implementing advanced analytics software, industrial organizations can easily corral data from numerous sources, automate its preparation for analysis and modeling and provide process experts with more time to study the insights and optimize operations for maximum uptime.
Learning Objectives
-
- Understand three obstacles to process industry data analysis and identify conventional software limits for process data analysis.
- See how advanced analytics software can automate data acquisition, accelerate insight generation and mitigate reactor failure, preventing unplanned downtime.
Process industry data analytic software insights
- Three obstacles can restrict process industry data analysis including conventional software limits.
- Advanced analytics software can automate data acquisition, accelerate insight generation and mitigate reactor failure, preventing unplanned downtime.
- Process industry data, even when diverse, can be analyzed centrally with ease.
The industrial internet of things (IIoT) and the Industry 4.0 emergence of artificial intelligence and machine learning (AI/ML) technologies massively multiplied the amount and complexity of data and shattered tried-and-true marching orders and business as usual for data gathering and analytics.
For decades, the process manufacturing industries had relied on a variety of systems for monitoring, gathering and processing data in real time, including supervisory control and data acquisition (SCADA), distributed control, laboratory information management systems and others. Procedures for data management and handling in operations technology (OT) environments were relatively straightforward and easy to follow for a large span of time.
After the Industry 4.0 data explosion, data management software slowly played catch up, providing engineering teams with greater visibility into historical and near-real time data from both local and remote locations. Despite this, many process manufacturers still struggle to smoothly translate raw data into meaningful insight. Why does it remain difficult for process manufacturers to gain insights from raw data?
Three obstacles to process industry data analysis
Process data bumps in the road most frequently result from:
-
Limited data access and connectivity tools.
-
A lack of time-series-specific analytical solutions.
-
Cross-functional team collaboration difficulties.
Addressing these issues is paramount for operational understanding and informed decision-making, in addition to progressing toward process optimization, sustainability and workforce empowerment corporate initiatives.
Conventional software limitations for process data analysis
In most facilities, numerous data sources exist, creating equipment, process, quality and inventory data, and this information was traditionally stored across multiple disparate databases. In these OT settings, spreadsheet-based analytics tools were used to access, cleanse and align all this data so insights could be extracted. This manual procedure was cumbersome and time-intensive for process experts, engineers and data scientists (collectively, subject matter experts, or SMEs), and it made inefficient use of these resources’ valuable skillsets. In fact, a 2016 CrowdFlower study found that SMEs spend nearly 80% of their data analytics time collecting and wrangling data into suitable formats for analysis, leaving only 20% to work toward meaningful insights (Figure 1).
In addition to these manual inefficiencies, the lack of live data connectivity among many facilities rendered SMEs’ analyses continuously out-of-date. These challenges made it difficult to wrangle and prepare data for meaningful analysis. Furthermore, sharing data and analyses across organizational teams and regions using conventional solutions was laborious or nearly impossible, limiting the ability for collaboration and knowledge transfer among peers and colleagues.
Spreadsheet-based applications are not optimized for time-series data analysis, and they exhibit poor visualization functionalities, prohibiting quick and iterative analysis. Although SMEs bring invaluable process knowledge and insight to the table regardless of the analytic software they use, traditional tools hinder the effective and efficient data analytics capabilities needed to improve production outcomes. Fortunately for process manufacturers, better solutions are now available.
Advanced analytics software automates data acquisition, accelerates insight generation
Advanced analytics solutions address these and other issues, connecting disparate data sources throughout enterprises to a central cloud-based or on-premises software system. These types of applications automate the cleansing and contextualization of data, while performing time-stamp alignment in the background so SMEs can quickly garner meaningful and transformational insights from all available data. Equipped with live data connections, these applications enable users to conduct analyses using near-real time data.
By breaking down data-access barriers, this automation empowers SMEs to leverage advanced analytics solutions’ purpose-built, time-series, analytical tools, typically in a no-code or low-code point-and-click format, to derive data insights. Coupled with often built-in trending and data visualization capabilities, users can visualize the impact of their data analyses to quickly iterate and make informed process improvements.
Advanced analytics software tools also enable process industry organizations to maximize the effectiveness of SMEs, who frequently work from different facilities or geographies, by enabling streamlined collaboration, knowledge capture and reporting. For example, product assets often change hands in the process industries, so it is common to find multiple sites within an organization that are struggling with the same issues.
Analytic software empowers organizations to implement enterprise-wide analytics strategies that promote cross-site collaboration, such as exchanging best practices for predicting and preventing common failure modes among assets. These analyses can then be shared and scaled across plants or product portfolios, and then results can be used to train new personnel.
Analytic software examples: Mitigating reactor failure, preventing unplanned downtime
Collaboration among and between teams, such as process, maintenance and reliability teams, can be strengthened especially by leveraging built-in tools within advanced analytics software for sharing analyses and insights in easily digestible dashboards and reports.
A petrochemical and refining company was experiencing significant reactor shutdowns caused by a failing critical feed gas compressor on a polyethylene line. There was no ability to immediately restart the process following such failures. At this facility, an unplanned reactor shutdown creates a minimum of four hours of downtime, costing the plant upward of $200,000 with each incident. Previously, these compressors had been maintained on a preventive maintenance (PM) schedule. The PM schedule did not prevent unplanned shutdowns.
Following one compressor trip, machinery and controls engineers worked together to identify the safety interlock that prompted the shutdown, bringing electrical engineers in to assist with the investigation. However, tracing electrical diagrams around the pump motor was time-consuming, and it failed to yield a root cause.
One process engineer at the refinery chose to take an alternative approach, opening a browser-based advanced analytics software, to rapidly locate the five most recent shutdowns and subsequent restarts, planned and unplanned, from decades of historical process data. Using time-dissection “Capsules” and “Chain View” tools, they quickly focused on the shutdown and startup time periods and overlaid all events, presenting abnormalities in the discharge pressure profile of the two most recent startups (Figure 2).
Upon further investigation, the engineer identified early warning signs on the motor amperage signal. Without a method to view the startups back-to-back, the motor degradation had gone unnoticed by operations.
As a result of the root-cause analysis, the process engineer implemented a monitoring solution to identify and flag future motor degradation to prevent similar unplanned shutdowns. When an out-of-tolerance value appears, the compressor motor is immediately added to the maintenance work list for the next planned shutdown, a proactive maintenance approach that is expected to eliminate unplanned shutdowns from this failure mode.
Handle diverse process industry data centrally with ease
Without comprehensive software solutions to connect to disparate data sources, provide intuitive tools for engineers and enable effective collaboration, key data analysis challenges will remain in process manufacturing organizations.
By empowering SMEs with modern analytic software, companies can make clear sense of and draw insight from their raw data, create analyses of varying complexity to model present and future behavior, and transition from reactive to predictive maintenance strategies. These outcomes are essential for industrial organizations to realize the full potential of their growing data repositories and to adapt to ever-evolving market conditions and needs.
Katie Pintar is a senior analytics engineer at Seeq Corp.; edited by Mark T. Hoske, content manager, Control Engineering, CFE Media and Technology, [email protected].
CONSIDER THIS
How can those with industrial automation expertise help their organizations take advantage of massive amounts of data and resolve data complexities?
ONLINE
More from Pintar and Control Engineering about continuous process improvement
