Solutions for Some of
Our Largest Partners
In six months we went from concept to proven.
- Brian Pugh, Chief Innovation Officer
Insight
Unsynchronized well arrival increases downtime
Event
AI-driven system-level queuing model
Action
Automated well arrival synchronization
Impact: Increased profitability and reduced environmental impact
• 20% production uplift
• 22% reduction in operating costs
• 74% reduction in methane emissions
We view the relationship with Kelvin as another step toward bringing industry-leading intelligence to the market.
- Michael Segura, Vice President of Production Enhancement at Halliburton
Insight
Uneven pump management increase service cost
Event
Real time load imbalance detection
Action
Automated load redistribution
Impact: Market leading capabilities deployed in under 90 days
• Multi-year R&D pipeline translated into production ready solution
• $10M reduction in annual operating expenses
Using Kelvin, our development teams can transform their ideas into production-ready algorithms faster than ever before.
- Helenio Gilabert, Senior Director, Digital Transformation Edge Solutions at Schneider
Insight
Unplanned pump failures causing downtime
Event
AI image classification of pump cycle cards
Action
Automated equipment failure prediction
Impact: Field scale, edge ML applications drive increased profitability
• 40% reduction in energy consumption
• 15% production uplift
• Increased operator efficiency
Kelvin Intelligent Control Software enabled easy system optimization, application deployment and ongoing model management while delivering productivity gains.
- Glenn Lydyard, Technology Operations (Digital)
Insight
Reliance on manual asset recovery decreasing asset performance
Event
Automatic detection of failure events
Action
Autonomous recovery from failure events
Impact: Reduced downtime and production uplift
• 8% production uplift
• 95% reduction in manual operator interventions
In a short amount of time we were able to prove the concept that operators and 'smart' software can work together to build reports and track key data, a bit step forward to our vision of reporting by exception only.
- Vittorio Spoldi, P. Eng, Well Performance Engineer, Shell
Insight
Reliance on manual workflows created poor data quality and low operator efficiency
Event
ML event detection on streaming data
Action
Real-time event driven workflow automation
Impact: Increased data quality and operator efficiency
• Idea to execution completed in under 30 days
• Generalized ML model deployed across all active jobs without re-training
• Seamless OT-IT integration