Leveraging AI/ML To Reduce Costs And Co2 Emissions

Optimize the electricity consumption

A leading Oil & Gas processing plant (one of the largest companies in the world) consumes large amounts of electricity to produce the gas needed to sustain its demand worldwide. Electricity cost is a key driver in their operational expenditure, and most of the consumption is caused by gas compressors. The status quo entails running these compressors at maximum throughput, which comes at a very high energy cost. However, maximum production may not necessarily be the best approach from an environmental and economical perspective. The client turned to Mobiquity to explore how it could optimize electricity consumption and reduce CO2 emissions while maintaining a high and stable production.

Artificial Intelligence & Machine Learning

To address this challenge Mobiquity applied Artificial Intelligence (AI) and Machine Learning (ML). We did this by leveraging our Digital Traction methodology, a framework for organizing digital innovation programs for optimal success. We provided this client with the following solution:

A Machine Learning model that accurately models the workings of the plant.

The model predicts how the plant will evolve over time and is able to accurately predict energy consumption.

By using this model, optimal settings can be discovered to optimize energy consumption and thereby reduce costs.

Digital Twin Solution

Together with Mobiquity’s data scientists, engineers, and technical and operational experts, the client was able to define project scope and success criteria. By leveraging a cross functional team, we were able to fast-track results and allow for quick iterations.

Part of our approach was to apply data science in a scrum-like fashion and enforce concrete results on a per-sprint basis. We achieved results by rapidly experimenting and creating an engine that enabled us to quickly train, optimize, and benchmark a large variety of AI/ML models. The Mobiquity outside-in approach to modelling focused on the end-to-end process first - before modelling the details.

Mobiquity created a data library, leveraging a practical digital twin that accurately mirrored the plant’s operations. By running a large number of AI/ML models, the system found the settings that would best optimize energy consumption and drastically reduce CO2 emissions. This innovative approach stems beyond human capacity, enabling the technology to conquer the groundwork for maximum efficiency.

Reduce the cost of electricity & CO2 emission

As a result of this project, the client has improved the quality of its electricity consumption forecast by 50%, significantly reducing the cost of electricity. We also helped to reduce the CO2 emissions in the amount equivalent to ~600 households in the UK.

With Mobiquity’s work, the client can also reduce the time to market for future AI/ML initiatives regarding operational optimization. Last but not least, this model has enabled the client to do remote maintenance investigation, which can result in reduced operational downtime, enhanced security, and increased productivity.

Leveraging AI/ML To Reduce Costs And Co2 Emissions

Opportunity

Optimize the electricity consumption

Services

Location

Share

Solution

Artificial Intelligence & Machine Learning

Digital Twin Solution

Results

Reduce the cost of electricity & CO2 emission

Bank of the Philippine Islands

Building a better digital banking experience for the Filipino people

Let’s talk about how to transform your friction into innovation.