Bridging the Physical and Digital Worlds Together in Energy

Historically, the approach to Operational Technology (OT), especially in environments like a refinery, was to keep the systems offline and completely disconnected from the Internet—known as the air gap—to prevent real danger if something went wrong. However, this isolation prevents the necessary flow of information, such as data about production volumes or maintenance routines, into AI and machine learning environments. The long-standing problem that Shell and Acuvate address is how to securely and reliably connect into the OT space over that air gap to get data in an integrated and accessible way. Shell is achieving this by trying to connect that OT environment into the office domain in an open source cloud first way to provide data visibility. 

Shell has developed an open source framework called the Real Time Data Ingestion Platform (RTDIP dot IO). This platform is designed to connect into real-time environments, such as those where data often resides in systems like OSIsoft (Aviva Pi). Shell has made RTDIP available through the Linux Foundation energy. This framework is part of the company’s efforts to use open source technologies for connectivity. Acuvate has embedded both RTDIP and its accompanying component, SSIP (SSIP PS), in its Acuvate Enterprise Data Platform for the energy and manufacturing spaces. 

The core function of RTDIP is to act as an extension to the lake house, specifically designed to handle time series data. It is crucial for picking up time series data and event data from OT. The purpose of RTDIP, alongside SSIP PS, is to ingest the data and store it in the Databricks environment. SSIP PS (a set of connectors built by Shell) works with RTDIP to help others adopt and embed the platform. 

While there is genuine movement regarding the adoption of standards like OPC UA and OPAF (Open Process Automation Forum), the adoption is slow. The reason for this slow pace is that many of the operational technology environments are quite static. The OT is often tightly coupled to the machinery behind it, meaning it is not easy to change and upgrade quickly, which holds back the rapid adoption of these new standards. This constraint means that the “digital speed often reflects the physical speed” of the asset upgrade cycles. 

The primary driver for connecting the OT environment and liberating the data is to enable the use of advanced analytical tools. Artificial Intelligence (AI) and Machine Learning (ML) specifically need to exploit this liberated data. For example, getting maintenance routines data, production volumes, and event data into the machine learning environment is necessary for enhanced operations. The availability of data allows for a “control tower” perspective, which tells operators where to go, which maintenance routines to do, and how much to invest, complementing the “pilot’s cockpit” view of the operator at the asset.