Optimize Industrial Maintenance with AI
Challenges
Every hour of downtime in a gas refinery costs millions of dollars. Maintenance planners at Petronas faced overwhelming complexity, shifting priorities, and conflicting goals with tools that left them reactive and relying on instinct.
Solution
I led the design of aiVA, an AI virtual assistant that provided planners with clear, actionable insights and real-time schedule recalibration.
Reduced task recalibration time by 50%
Gave gas refinery operators more confidence
Shift focus from chaos to proactive decision-making
Company
Aitomatic
Position
Founding Designer
Year
2023
Market
Asia, B2B
Challenges
Every hour of downtime in a gas refinery costs millions of dollars. Maintenance planners at Petronas are tasked with coordinating hundreds of interdependent tasks under constant pressure to keep operations running smoothly. Their job is a balancing act—responding to sudden defects, shifting priorities, and conflicting goals, all while trying to avoid costly delays.
The tools they used didn’t make things easier. Instead of providing clear guidance, these tools often left planners overwhelmed by chaotic workflows and endless recalculations. Decisions were made reactively, relying on instinct rather than reliable insights. The stakes were high, the pressure relentless, and even small mistakes could have massive consequences.
Solution
To tackle these challenges, we designed aiVA, an AI-powered virtual assistant built to bring clarity and control to high-stakes maintenance planning.
I led the design of aiVA, working closely with planners and engineers at Petronas to ensure the platform met their needs. The first step was to understand their workflows. Through observations and interviews, I saw how planners struggled to trust tools that gave recommendations without explanations. They needed transparency—clear reasoning behind every suggestion. They also needed flexibility, with a system that could adapt instantly when priorities shifted.
aiVA was designed with these insights. It didn’t just rank tasks—it explained why each one mattered. For example, if a pipeline defect was flagged, aiVA highlighted the urgency, outlined risks, and suggested next steps to resolve the issue. This level of detail gave planners the confidence to trust its recommendations.
When priorities changed, aiVA recalibrated schedules in real time, ensuring tasks stayed in sync without overwhelming planners. It also turned engineers’ years of experience into a shared resource, creating a knowledge base that made expertise accessible to everyone on the team.
To ensure aiVA worked seamlessly in Petronas’ high-pressure environment, we tested prototypes through workshops and real-world simulations. Planners provided feedback that shaped every iteration, making the platform intuitive and practical. Early results showed that aiVA cut task recalibration time by half, freeing up planners to focus on preventing downtime rather than reacting to it.
Impact
aiVA fundamentally changed the way maintenance planning worked at Petronas. Tasks that previously took hours to adjust were now done in minutes. Planners trusted the system not just to organize tasks but to guide them toward the right decisions with clear explanations. This trust allowed them to shift their focus from managing chaos to improving operations.
Although still in the pilot phase, aiVA’s potential was huge. Petronas could expand its use across more refineries where aiVA enables smarter, more proactive processes. As one planner said, “aiVA gave us clarity in situations where we used to feel overwhelmed. It’s a partner that helps us stay in control.”
