OT/ICS Cybersecurity

A couple of years ago, Larry O'Brien from ARC Advisory Group wrote a paper titled ‘Sensor Data Integrity for Process Plants is a Problem: Here’s How Users Can Solve It.’ This issue remains unresolved for most customers, and I frequently engage in discussions with Owner/Operators in heavy processing industries on this topic. I believe it's time to bring attention back to this paper.  

Although the paper primarily targets Owner/Operators, similar conversations are arising with Engineering Procurement Construction (EPC) companies, and Main Automation Contractors (MACs). These discussions focus on data integrity across multiple systems, though the implications differ from operational facilities. EPCs and MACs are focused on minimizing data mismatches during Factory Acceptance Testing (FAT), Site Acceptance Testing (SAT) and more importantly, startup phases.  

For instance, we did a study with a greenfield facility that revealed significant inconsistencies between the instrument database, DCS, PLC and process historian post-FAT, with approximately 6% of instruments still having configuration discrepancies. They estimated a cost ratio of 10 to 1 between rectifying issues during startup versus addressing them in the factory setting. This illustrates that startup fixes can be 10x more expensive. This example resulted in substantial cost savings amounting to millions of dollars. 

While Larry's paper primarily addresses operational challenges for Owner/Operators, the issues faced by EPCs and MACs are similar but contextualized differently. This blog serves as a concise summary of Larry’s insights. 

Introduction to Larry’s Paper 

The evolution of intelligent sensors in the process industries has brought both promise and complexity. These sensors, embedded in field devices, control valves and analytical equipment, were expected to revolutionize device management and maintenance. However, they have also introduced significant data management issues, particularly concerning data integrity. Sensor data often show inconsistencies across various applications, resulting in no unified version of the truth regarding the sensor's state. This inconsistency impacts change management, device maintenance, plant performance and cybersecurity, contributing to unplanned downtime that costs the process industries significant losses annually. 

The Root of Data Integrity Problems 

Intelligent devices, which constitute over two-thirds of installed process field devices, have onboard microprocessors and bidirectional communication capabilities. While these features enhance device reliability and availability, they also create vast data inconsistencies. For example, applications like Hexagon's Intergraph Smart® Instrumentation are used to capture baseline configurations during system design. However, during installation, commissioning and routine maintenance, configurations may change, and these changes are often not logged properly, leading to discrepancies across systems such as Plant Asset Management (PAM), Distributed Control Systems (DCS), and Safety Instrumented Systems (SIS). 

The Importance of Sensor-Level Data Integrity 

Beyond data management, the issue is fundamentally one of data integrity. Inaccurate records of installed devices, firmware versions and change logs can escalate into major operational problems. This lack of visibility and control creates cybersecurity risks, as poor asset inventory management can leave vulnerabilities unaddressed. With the increasing number of sensors in process plants, driven by the deployment of IoT and wireless sensors, the scope and complexity of maintaining data integrity grows exponentially. 

The Financial Impact of Poor Data Integrity 

Unplanned downtime, often due to sensor or valve failure, is a significant cost for the process industries. Accurate sensor data is crucial for maintaining plant performance and preventing unexpected shutdowns. The failure of critical field devices, such as control valves, often leads to these shutdowns, with unplanned downtime costing industries more than $1 trillion annually per the paper referenced above. 

PAS Automation Integrity™ - Sensor Data Integrity Module 

Manual methods to tackle data integrity problems, such as using spreadsheets, are time-consuming and error prone. Hexagon’s PAS Automation Integrity™, leveraging the Sensor Data Integrity module, offers a vendor-neutral approach to managing and ensuring consistent sensor data integrity across multiple applications and systems. This solution, part of the PAS Automation Integrity suite, centralizes visibility and configuration management for both traditional and IoT-connected sensors. It facilitates new sensor discovery, configuration management and error detection, automating tasks that were previously manual and prone to errors. 

Benefits of the PAS Sensor Data Integrity Solution 

Case Study: Major Oil & Gas End User 

A major oil & gas company using Hexagon’s Sensor Data Integrity module has streamlined its processes and improved data consistency across DCS, SIS and PAM systems. Previously reliant on manual data validation, the company now benefits from automated data management, reducing the burden on engineering and maintenance teams and ensuring consistent device configurations. 

Conclusion 

The proliferation of intelligent and IoT-connected sensors in process plants will only exacerbate data integrity challenges. End users must move away from manual data management processes and adopt solutions like Hexagon’s Sensor Data Integrity. This vendor-independent solution provides comprehensive sensor data management across multiple applications and systems, ensuring operational efficiency, enhanced reliability and reduced cybersecurity risks.