Model the infrastructure of an oil and gas plant to enable predictive analytics for maintenance optimization. Pinpointing problems and suggesting solutions.


Model infrastructure at an existing oil and gas plant to capture critical operational, performance and environmental data in order to understand existing maintenance patterns. Use this information to develop a predictive analytics engine that can be used to power a preventative maintenance programme.

Insight and Action

In April 2020, Takamul began working on the development of a data platform based on Process Hub, a cloud-based solution developed for the oil and gas industry. The platform was connected to live data streams of process and sensor data to build up an understanding of current operational performance and how this interplays with points of failure and existing reactive and proactive maintenance activities. This understanding was then used by Takamul data engineers to develop the AI model for predictive analytics. Within a relatively short time, a digital model of the plant was developed. This digital twin behaved like the real oil and gas plant and was programmed to react to changing parameters in a realistic way.


The BETA environment was delivered in August with process engineers testing the models. The project has been extremely collaborative, with knowledge validation on both sides. As the data team simulated real-world scenarios, Aramco engineers worked alongside to validate models. Work is now ongoing to extend the timeframe of predictions, so that notifications about potential failures or maintenance requirements arise even earlier in the process. Wider deployment is scheduled for November 2020.


The King Abdulaziz Center for World Cultures

Read Use Case