Strategic application of AI can help save costs and bring efficiencies

The digitalisation drive in recent years has led to the deployment of technologies such as Artificial Intelligence (AI) and Machine Learning (ML), ushering in a new era of Industry 4.0. According to Gartner , 37 per cent of organizations have implemented AI in some form with the percentage of use cases growing by over 270 per cent in the past four years. Some immediate benefits have ranged from automation, and enhanced customer experience, to smart decision-making leading to better productivity, efficiency, and safety. Today, its use in traditional capital-intensive industries, such as oil and gas, is bringing rich dividends to the sector and the environment.

While AI helps enhance customer experience and supports smart decision-making by drawing on data insights, its impact in helping save capital expenditure, especially in manufacturing, is less discussed. With the proliferation of data generated by the Internet of Things (IoT), AI/ML has tremendous potential to deliver predictive insights and significantly curtail capital expenditure. However, opportunities remain as many industries are yet to fully tap into the potential of these newer technologies.

Artificial learning for smart outcomes

Today, most heavy industries have embarked on their digital transformation journey, which has helped unlock large data volumes regarding operational efficiency, productivity, and recurrent costs. Despite initial benefits, many organizations will soon face a challenge in supporting their ageing infrastructure with highly capital-intensive instrumentation and sensors. The key to optimizing sensor usage is to select vital instruments rather than applying them universally. Real-time monitoring is ideal with sensors on every instrument, but cost constraints dictate the careful selection of critical instruments. Despite being cost-effective, this approach does not allow for monitoring and controlling all instruments.

AI/ML uses data-learning algorithms to analyse hidden patterns and accurately predict results in real-time through the Internet of Things (IoT) and the Cloud. The results derived from this small sample size can be applied across remaining assets, with confidence, helping companies limit the cost of installing expensive sensors and other infrastructure equipment.

Updating heavy industries

To illustrate this further, consider an upstream oil & gas company that needs to inject polymers of the right viscosity in its oil wells. To measure the viscosity at each wellhead, it needs to install viscosity meters which are very costly. As an alternative to placing sensors on all the wellheads, which may go up from 100 to 1,000, the company can use AI/ML to analyse the viscosity of the polymers by installing sensors in just 15 to 20 per cent of wellheads. For the balance of 80 per cent wellheads, the company can use the result of AI/ML prediction coming from installed sensors and, thus, avoid the cost of installing extra instrumentation and sensors.

Another example is the usage of AI in enhancing the analysis of seismic data for exploration and reservoir characterization. Before using AI, one needs to ensure that there is enough computational capability to run the basic AI engine.

Cairn’s Mangala field has emerged as a gold standard in digital innovation for the Indian oil & gas sector - optimizing artificial lift systems, automating reservoir and production management, and enhancing boiler and STG efficiency through advanced algorithms. A holistic approach to asset management reduces the carbon footprint of the company, while AI-based surveillance ensures employee safety – demonstrating the company's ability to elevate productivity, sustainability, and safety.

At Cairn, as a part of our Integrated Petrotechnical project, we lifted and shifted complete geoscience and reservoir engineering software technologies which accounted for nearly 50 application suites across the E&P domain from 25 different technology partners. Along with the applications, more than 550 TB of data, backup and files, several databases and servers were also migrated. This has now given us the power of the Cloud to run AI/ML models. Now, by processing vast amounts of seismic data quickly and accurately, AI can help geoscientists identify potential drilling sites with higher precision, reducing the risk of costly dry wells.

In this way, AI/ML is assisting the company in conducting accurate simulation exercises, thus, saving money on capital expenditures. This process can be replicated for almost all industries which require heavy capex investments, such as telecom, power and energy, etc. The insights that machine learning provides can not only curtail capex but also help cut downtime, reduce costs, and increase productivity.

Further, for telecom companies that are expecting high-value capital investments in infrastructure with the deployment of 5G and edge-cloud networks, AI/ML can help in reducing capex costs. Network expenses such as tower installation, laying of fibre, etc., represent 25-35% of the operator’s cost structure – and AI/ML can significantly impact the bottom line by automating tasks and improving network plan and design. This improvement will ultimately enable a more efficient capex per site spending by creating a leaner network grid.

The utility sector too can benefit from AI/ML by predicting consumption levels in specific locations by juxtaposing household, weather, and historical data to decide the optimal capacity levels needed to avoid outages. This will help in critical decision-making regarding how much capital infrastructure is needed for a particular area.

With organizations increasingly focusing on cost optimization, AI/ML will gain traction in the near future to manage capital expenditures.