Radical OT Data Transformation: Enhancing Excellence with Four Abilities

Scarce Data Owing to Data Silos

One of the impediments to data integrity is the scarcity of data. This shortage primarily stems from the fact that automation systems weren’t architected with data scrutiny in mind. Even in instances of data transfers across the factory floor, data is usually extracted to bolster operational equipment management exclusively, which falls short in delivering actionable business insights. For instance, consider a manufacturing plant housing a critical machine on its production line that is pivotal for all manufacturing operations. When this machine malfunctions, the entire production line grinds to a halt. Predicting potential failures in the machine’s crucial components to minimize downtime necessitates accessing data that the devices seldom provide about their integral parts. Consequently, installing sensors in these components and transforming the analog signals they emit to digital signals using remote I/Os becomes imperative. These digitized signals can then be relayed to upper-tier servers or cloud platforms to enable predictive maintenance. This showcases the prowess of OT Data Collection.

In the mentioned scenario, dealing with a single machine is a manageable task. However, when confronted with an entire factory equipped with multiple communication protocols, the complexity of conversion escalates significantly. Given that OT systems have operational spans of several decades or more, machinery from diverse vendors is often intermingled within the same ecosystem. Moreover, each equipment boasts its proprietary hardware blueprint, communication interface, and communication protocol to assure OT-grade reliability. This strategy effectively safeguards system uptime and optimizes performance if the systems operate autonomously. Nonetheless, over time, data silos have evolved. When endeavoring to consolidate data from distinct systems, a factory encounters the challenge of diverse communication protocols, akin to each system communicating in its unique dialect. For instance, two production lines within the same facility may deploy disparate PLCs from distinct manufacturers, each deploying its exclusive communication protocol for the respective PLC.

Thankfully, the market has recognized this conundrum. Numerous remedies are now available, including the adoption of standardized and open interfaces like OPC-UA or industrial protocol gateways, facilitating the extraction of machine data utilizing familiar communication protocols. For instance, leveraging Modbus-to-BACnet industrial protocol gateways, an HVAC system can ingest Modbus RTU data via the BACnet protocol.

IrrelevantData Lacks Utility for IT

Another hurdle obstructing data integrity is the presence of inscrutable data. Data generated by machinery are essentially raw values or metrics. IT analysts are unable to harness this raw data directly, with manual intervention impeding real-time responses. Converting OT data into meaningful values initially enables seamless and swift data flow across the edge-to-cloud architecture. OT data comprises a series of time-stamped numerical values representing specific events pertaining to particular devices or sensors at precise instants, like the instantaneous power draw of a motor every 10 seconds over the past week. On the contrary, IT data is structured data residing in databases, featuring meticulous structures and descriptors, necessitating contextual enrichment before being employable for diverse analyses. Of the earlier-mentioned OT data, the information displayed is restricted to the numbers 7 and 10; hence, preprocessing is crucial to infuse complete meaning (timestamps, units, etc.) into the data by appending the contextual details. Only then can comprehensive analyses be conducted.

Furthermore, for maintaining optimal control and effaccuracy, Operational Technology (OT) gear frequently generates a data fragment once per second or every millisecond. If all the unprocessed OT data gets sent to an Information Technology (IT) system, it may overwhelm the system’s capabilities. Even worse, transmitting irrelevant data to cloud services not only hampers operational efficiency but also escalates data transfer and storage expenses. To address this issue, intelligent Internet of Things (IoT) devices are employed to manage the frequency of data dissemination. This allows OT equipment to synchronize with IT requirements, such as uploading data once per hour or analyzing data on the OT side before uploading it when significant deviations are noticed. These measures are crucial for excelling in OT Data Preparation.

Diverse Origins Leading to Incomplete Data

The ongoing digital transformation necessitates a broader range of real-time data, resulting in an increased flow of OT data. While traditional OT networks transmit data to meet control mandates, industrial digital transformation demands data transmission for analysis and decision-making purposes. Consider a smart factory, for instance. To achieve flawless production, assembly lines must provide immediate feedback at each stage. When an irregularity is detected – indicating an issue at a previous station – the consequent station will promptly alert the former station to trigger an instant reset, preventing minor deviations from accumulating and causing failures. Consequently, vast volumes of data need to traverse OT networks, encompassing control details and defect illustrations. Concurrently, a new challenge arises: How to prevent impeding OT control data transmission with the integration of IT data?

Why is this an issue? It’s due to the absence of real-time control mechanisms for bulk data in industrial Ethernet networks, the most widely used networks. The proposed remedy involves the use of two segregated networks for transmitting images and control inputs. This ensures that both streams of data don’t vie for network bandwidth, although it results in doubled costs for network setup and maintenance. Time-sensitive networking (TSN), the latest Ethernet iteration, is engineered to schedule transmissions based on data importance, guaranteeing that critical data reaches the device as scheduled. This is the quintessence of robust OT Data Transmission capability.

Furthermore, various environmental interferences, like extreme temperatures and electromagnetic interference during device startup, can disrupt networks and lead to data loss en route. Consequently, backup plans should be instituted to mitigate incidents and prevent data loss due to interruptions during transit. For example, in the event of network failure, a network backup mechanism can swiftly initiate another segment to resume transmissions. Or, in scenarios of network congestion or disconnection, the most recent data can be stored locally to ensure that if lost, it can be either retransmitted or retrieved to prevent fragmented data delivery.

Data Vulnerability Stemming from Security Flaws

OT data reliability is often compromised due to cybersecurity vulnerabilities. Historically, OT systems maintained an air-gap from the internet, achieving protection through physical control means like access restriction to OT regions or prohibition of portable drives and personal computers. However, with industrial digitalization, internet connectivity becomes inevitable. Consequently, all vulnerabilities become exposed to merciless computer viruses or a tempting target for profit-driven hackers, offering entry points to infiltrate systems and potentially disrupt operations. As cyber threats proliferate, data security and cybersecurity have emerged as essential components of every digital transformation agenda. To safeguard production capability and immunize production lines against data manipulation assaults, organizations mustn’t allow their operations to fall prey to the Achilles’ heel of untrustworthy OT data.

A prevailing misconception among businesses is assuming that mature IT security solutions can seamlessly transition to their OT realm; however, security tools designed for IT environments may not be entirely suitable for OT protection. For instance, antivirus software is ubiquitous in the IT world, but OT devices don’t necessarily run compatible operating systems to support antivirus installations. Moreover, the preeminence OT environments assign to operational continuity deters many from deploying antivirus solutions due to fears of hindering production capacity by erroneously blocking data packets. Furthermore, to ensure connection stability and convenience, manufacturers might interconnect all devices on a singular intranet. Nonetheless, once ransomware infiltrates, it can swiftly propagate across the entire system. Hence, securing OT environments via three progressive stages is recommended: endpoint security, cybersecurity, and security management. To bolster OT Data Security capabilities, enterprises should:

• Implement Intrusion Protection System (IPS) technology on OT automation devices to safeguard critical infrastructure. An industrial-grade IPS vigilantly monitors data inbound and outbound of critical devices, segregates suspicious traffic, and promptly alerts administrators upon anomaly detection.
• Leverage network segmentation to mitigate ransomware incidents. Companies can upgrade their Ethernet switches to managed counterparts and activate segmentation features to split an OT network into isolated sections.
• Employ network management software to overcome interoperability issues among various OT communication protocols, facilitating efficient identification of flawed or insecure devices through visualization.

Harnessing the Power of Four OT Data Capabilities

As the saying goes, “don’t get ahead of yourself”. It is imperative to align your priorities during industrial digital transformation. Do not let poor-quality raw data compromise the efficacy of your big data analyses. Before sourcing OT data, contemplate your status in terms of data acquisition, data preparation, data transmission, and data security. Equipped with these four capabilities, you will be well-equipped to confront challenges head-on and leverage high-quality OT data to establish a robust foundation for transformation.

Do you wish to delve deeper into the intricacies of OT data?​​ Listen to OT Data Next here: