In my conversations with energy leaders, one challenge comes up again and again. They have more visibility into their operations than ever before, but that doesn’t necessarily make the decisions any easier.
For energy operators across Asia, the stakes are particularly high. Geopolitical shifts can quickly affect feedstock availability, pricing, supply chains and energy costs. Operators must manage that volatility while maintaining assets and meeting growing demand for reliable and affordable energy. That demands both resilience and agility.
Operational excellence increasingly depends on how quickly companies can interpret those shifts and respond with confidence. But as the number of possible scenarios grows, making the right call becomes that much harder.
When the choices multiply
Historically, teams could manually analyse a limited number of scenarios. That worked under more predictable conditions. Today, the possibilities can quickly outstrip what teams can assess manually.
Think about what happens when one variable changes. A shift in feedstock availability may affect costs, production plans or maintenance decisions, creating another set of possible outcomes. Multiply those variables across an entire operation, and complexity quickly compounds.
This is where industrial artificial intelligence (AI) earns its place. It can evaluate far more scenarios than teams could analyse on their own. Instead of narrowing the choices too early, companies can weigh the trade-offs before making the call.
AI doesn’t make the call
That said, I’m not suggesting we simply hand the decision over to AI. The scenarios it helps us to explore still have to make engineering sense, grounded in first principles and the laws of physics that govern how a plant actually operates.
AI is extremely good at finding patterns in data and doing much of the analytical heavy lifting. But people bring operational experience and engineering judgement. They define the objectives, constraints and desired outcomes. AI takes on the analysis and scenario modelling.
This is augmented decision-making, not automated decision-making. AI expands what people can analyse. People still make the call.
What matters is how much more they can understand before they do.
From visibility to anticipation
The first generation of digital twins gave companies a snapshot of a process or plant. Living digital twins go further. They combine real-time data, operational context and AI to learn how a system behaves and how its different parts influence one another. That takes the digital twin beyond observing operations towards anticipating how they might respond to change.
At AVEVA, we are taking this further with AI-assisted twin building and industrial knowledge graphs to give AI richer operational context. By connecting information across engineering and operations, teams can better understand what is changing and what it could mean.
Instead of only asking what is happening now, teams can explore how different conditions might play out. What if production is adjusted? What if an asset starts behaving differently? How might a change improve efficiency without affecting emissions?
Rehearsing decisions before putting them into practice can strengthen operational resilience.
Thailand offers a practical example. At SCG Chemicals, an AI reliability platform built around a digital twin helps teams detect process deviations early and prepare for emergencies. The platform has delivered 99% plant reliability, and a ninefold return on investment in just six months.
From operational excellence to decision excellence
For decades, operational excellence has been measured through reliability, throughput, utilisation and efficiency. Those measures still matter. But as complexity grows, better decision-making is becoming just as critical.
That is what I look forward to discussing with energy leaders at Gastech in Bangkok. Greater foresight can help us build more resilient operations.
The hard calls are not going away. What can change is the quality of judgement that informs them.
Leave a Reply