United KingdomEnergy and utilities2018 – ongoingShell
Predictive maintenance on the C3 AI platform
Shell scaled machine-learning failure prediction from a pilot to more than 13,000 pieces of equipment
Problem
Unexpected failures of control valves, pumps, compressors and other critical equipment at Shell's production, refining and gas sites cause unplanned downtime and lost output.
Approach
Starting in 2018, Shell worked with C3 AI to build machine-learning models that read sensor data and flag equipment likely to fail, so repairs can be planned before a breakdown. The program reached 10,000 pieces of equipment across upstream, manufacturing and integrated gas assets by early 2022, and in June 2026 Shell signed a new multi-year agreement to add agent-based root-cause analysis.
Results
Takeaway
The programme grew from a pilot to a portfolio-wide system over eight years with the same supplier. All scale figures came from C3 AI announcements quoting Shell executives; C3 AI claimed "hundreds of millions of dollars" in value, while Shell published no savings figure of its own.
Categories
- Size
- Large enterprise
- Industry
- Energy and utilities
- Business function
- Production and quality
- Innovation type
- Process innovation
- AI technology
- Predictive analytics and ML
- Adoption
- Co-developed with a partner
- Evidence
- Vendor's case study
Sources
Source summaries
C3 AI announces that Shell's predictive maintenance program reached more than 10,000 pieces of equipment across upstream, manufacturing and integrated gas assets, a target Shell had set for 2021. It describes monitoring of control valves, pumps and compressors to avoid unplanned downtime, and quotes Shell's Dan Jeavons on reaching the target.
The trade journal reports the 10,000-equipment milestone and gives C3 AI's figures for the system: 20 billion rows of data a week from more than 3 million sensors, and nearly 11,000 models making over 15 million predictions a day. It notes a five-year renewal of the Shell–C3 AI agreement in 2021 and that no savings figure was disclosed.
C3 AI announces an expanded collaboration with Shell after five years of work together. Shell's Dan Jeavons is quoted saying more than 10,000 pieces of equipment from tens of Shell assets are being monitored and that the work helps reduce cost and improve reliability. No savings figure is given.
IT Brief reports a new multi-year agreement under which Shell extends C3 AI Reliability across its global assets and adds agent-based root-cause analysis. It says the relationship began in 2018 and the program monitors more than 13,000 pieces of equipment, and repeats C3 AI's claim of "hundreds of millions of dollars" in economic value.
Full details are in the original sources.
