United KingdomEnergy and utilities2025 – ongoingSeadrill
AI-based condition estimates for blowout preventer parts within asset lifecycle management
Seadrill used machine learning to estimate the remaining life of blowout preventer parts on 11 drillships
Problem
Blowout preventer (BOP) parts such as annular elements were inspected and overhauled on fixed intervals. Seadrill wanted maintenance decisions based on the actual condition of each part.
Approach
With ADC Energy, Seadrill built Annu-Life, a platform that combines rig data with machine learning models checked by ADC engineers and a database of images of rubber degradation. A supervised-learning image model grades annular element damage from photos on a scale of 1 to 5. Models for ram elements were in field testing. Most predictive functions in Seadrill's wider lifecycle platform still rely on rule-based thresholds because contractor data is limited.
Results
Takeaway
Seadrill paired machine learning with rule-based monitoring and said data limits held back the ML part. The 50% cut belongs to the whole lifecycle programme, not to AI alone.
Categories
- Size
- Large enterprise
- Industry
- Energy and utilities
- Business function
- Production and quality
- Innovation type
- Process innovation
- AI technology
- Predictive analytics and MLComputer vision
- Adoption
- Partner co-development
- Evidence
- Company's own claim
Sources
Source summaries
News item on the Annu-Life platform from ADC Energy and Seadrill, which estimates the remaining useful life of BOP components. It says the platform ran on 11 drillships and that extension to ram elements was planned. It gives no performance metrics.
Article by a Seadrill author on its asset lifecycle management platform. It describes image grading of annular elements, ram element models in testing and mostly rule-based predictive functions. It reports a 50% cut in overhaul and inspection frequency from the programme.
Full details are in the original sources.