Webinar | Energy Case Studies 2

The energy sector is generating more data than ever - however transforming that data into meaningful outcomes remains a key challenge. Join this webinar hosted by Technology Catalogue and discover how VROC’s AI platform empowers oil and gas teams to apply machine learning at scale to rapidly solve a variety of operational problems.

Hear directly from Razik M Aznam, of PETRONAS Carigali (Turkmenistan), share how their offshore team utilises AI to continuously monitor processes and equipment. He will share an example where AI models predicted a failure with a Gas Turbine Generator, which allowed their team to plan corrective measures and prevent a trip.

About the Speakers

Trevor Bloch is the Chief Technology and Products Officer of VROC, is the driving force behind this pioneering company. With a background rooted in traditional consulting control system engineering, Trevor has leveraged his extensive industry expertise to establish VROC as a powerhouse in the field of industrial information technology.  Under Trevor's leadership, VROC has grown into a powerful end-to-end data and AI platform that adds substantial value to clients by harnessing time series data and breaking down information silos, enabling industrial AI to be implemented at scale. 

Razik Aznam is a seasoned professional whose extensive experience spans 20 years in the oil and gas industry. Over the course of his career, he has held progressive roles including Head of Production Operations and Chief Operating Officer PETRONAS Carigali, where he led strategic initiatives, optimized processes, and delivered measurable results.

• Predictive AI delivered material reliability gains, including multi-day advanced warnings of equipment failure, elimination of unplanned generator trips, and >USD 20M in loss avoidance across upstream assets.
   
• Automated machine learning uncovered non-obvious root causes—often outside the equipment boundary—compressing diagnostics from weeks of engineering effort to minutes.
   
• Centralized, portfolio-level monitoring enabled proactive prioritization of high-impact interventions, directly aligning maintenance execution with production and HSE KPIs.
   
• Successful adoption required cultural alignment and targeted upskilling, shifting engineers from manual troubleshooting to data-driven, high-value decision support.

The webinar, jointly hosted by TechnologyCatalogue.com and VROC, focused on translating industrial data into operational impact through real-world applications of machine learning and predictive analytics. The session showcased how upstream operators are leveraging VROC’s AI/ML platform to drive asset reliability, optimize production, and compress decision-making cycles across complex facilities.

Erik Nijveld, CEO and Co-Founder of TechnologyCatalogue.com, opened the session with context on the rising demand for scalable, low-friction digital solutions that materially influence uptime and cost performance. He then introduced the presenters: Razik Aznam, formerly Chief Operating Officer at Petronas Carigali Turkmenistan, and Trevor Bloch, VROC’s Chief Technology & Product Officer.

Real-World Use Cases from Upstream Operations

Razik outlined two field deployments where VROC’s automated machine learning played a decisive role in performance improvement.

1. Gas Compressor Reliability Issues (Malaysia)
 Operating a critical gas-lift dependent crude platform, Razik inherited a gas compressor with mean time between failures of roughly two weeks. Conventional troubleshooting— utilising OEM support, expert mobilization, and repeated root cause cycles—did not resolve the chronic reliability issues. By streaming platform-wide historian data into VROC, the team surfaced 20+ previously overlooked causal factors within minutes. Contrary to long-held assumptions, the primary reliability driver was not within the rotating equipment itself but in upstream gas handling conditions. Focused interventions subsequently increased monthly operating hours significantly, reducing unplanned outages and generating an estimated USD 21.7 million loss avoidance at the time.

2. Generator Failure Prediction (Turkmenistan)
 A second deployment centered on turbine generator availability. System philosophy required two units online with one in standby, limiting flexibility during extended outages. Using VROC’s predictive models, the team proactively identified a failure mode driven by the air-intake filter differential pressure. The platform forecasted failure 5.6 days in advance, enabling controlled shutdowns, load redistribution, maintenance execution, and restart sequencing without production impact. The model provided clear probability-weighted root causes and tracked deterioration patterns, allowing the team to prevent multiple unplanned trips. This predictive capability ensured uninterrupted power supply for the entire complex—an operational priority given the high-dependency process configuration.

3. Loss of Primary Containment (LOPC) Avoidance Through Anomaly Detection
 Razik also showcased a case where the platform’s moisture analyzer persistently triggered high readings, placing the facility at risk of mandatory shutdown under corporate corrosion-control policy. VROC’s model pinpointed an unexpected driver: a specific gas-lift well whose malfunctioning valve system caused recirculation and elevated moisture levels. Although counterintuitive to operators, shutting in the well halved moisture levels immediately. The intervention prevented a full-platform shutdown, generating USD 1.12 million in loss avoidance.

Across all cases, Razik highlighted not only the technical impact but also the organizational shift required. Successful adoption depended on building internal trust, upskilling engineers to interpret AI-generated insights, and repositioning human expertise from manual investigation to targeted action.

Technology Overview and Strategic Direction