I have my head in the clouds

The title has two meanings for me. Firstly, I literally have my head in the clouds as I have been exploring cloud computing. Secondly, I feel like I am living in a fantasy world where I have access to unlimited computing resources and domain expertise from any part of the world.

Just like clouds in the sky, cloud computing is ubiquitous and versatile. There are different ways to use cloud services: Software as a Service (SaaS), Platform as a Service (PaaS) and Infrastructure as a Service (IaaS).

In a recent application of cloud computing, I used IaaS to migrate on-premise SCADA servers to Microsoft Azure. These servers collect data from offshore platforms using satellite connections provided by a third-party vendor. The primary reason for this migration was scalability; as demand from clients increased at an unprecedented rate, the existing infrastructure reached its capacity prematurely. In a traditional on-premises approach, capacity increase would require capital expenditure (CapEx) approvals. This highlights the second compelling reason: shifting investment from large CapEx to smaller operational expenditure (OpEx). The justification of CapEx is more challenging than OpEx because it requires a forecast of future demand over the next two to three years. This forecast period is usually related to the typical lifespan of infrastructure — which becomes irrelevant in cloud computing. Incidentally, with cloud computing users are released from the tedious tasks of maintaining on-premise hardware.

During the proof of concept for this migration, I was on cloud nine because I could execute all of the tasks from the comfort of my office. These tasks included testing connectivity from a remote terminal unit (RTU) to a SCADA server in Azure, collaborating with the satellite connection vendor located 3000 km away and interacting with end users to optimise the user experience at their own workstation.

The elastic and centralised nature of cloud computing could come at the expense of data reliability, latency, network bandwidth and data security if we are not careful. Therefore, a robust security architecture is vital. Security should be built into data generation at the devices, data ingestion at the edge, data transport between edge and cloud, and data protection in the cloud.

However, even the most robust and well-designed security architectures can be undone by poor practices. To reinforce the vulnerable link between technology and real people, enterprises need to support users with proper training and a strong cultural awareness of data security.

With security architecture and competent users in place, IT can apply existing infrastructure to detect anomalies and remove potential threats. Besides being the guardian of the enterprise’s cyber attack surfaces, IT also ensures proper network segregation within the enterprise and harmonisation between different business units, with OT being one of them. Consequently, OT should work with IT to ensure success in their IIoT cloud deployment.

As I see it, the IIoT cloud unlocks plenty of possibilities for the OT world. To rise above the dark clouds of improper deployment, enterprises should arm users with best practices in data security and advocate IT/OT collaboration.

Mei Wan is a chemical engineer who has spent 15 years delivering process automation projects for a range of industries. As the Solutions Architect for Emerson Automation Solutions in ANZ, Mei helps clients tackle business challenges with IIoT and other digital technologies.

Image: ©stock.adobe.com/au/T. L. Furrer