While the advantages are innumerable, pockets of doubt and cynicism within organizations can limit the successful outcomes of adopting AI.

In order to extract actionable insights out of data, organizational and IT leaders need a mindset shift.

Everyone in an organization has to have an understanding of the benefits and the processes of building intelligence capabilities based on a data-driven approach. Also, employees must be given the opportunity to understand how these technologies can be applied to tackle challenges through first-hand experience coupled with digital skills training programs designed not only for building data scientists but also intelligence analysts and AI strategists in the organization.

This is the opinion of Luis Gonzalez, GM for Data Innovations in the Power Industry, Aboitiz Data Innovation (Philippines), who kindly responded to DigiconAsia’s request for his insights on optimizing AI adoption.

DigiconAsia: During the pandemic, businesses may not have the means to adopt AI and machine learning tech over basic digital transformation measures.

Luis Gonzalez (LG): Nowadays, with open-source models and libraries, and free cloud computing access, all it takes is creativity and innovative mindset to begin. The opportunistic approach of AI requires pragmatism in how to innovate with the existing data as well as simplify the outcomes required.

DigiconAsia: Can you give examples of AI being deployed to reduce costs, streamline processes and generally make life easier for people?

LG: One example of using computer vision in manufacturing lines to speed up quality assurance.

Another use case of data science and AI is helping the unbanked and under-banked with alternative scoring systems to assess their eligibility and provide greater access to financial support and services. In one example, the Union Bank of the Philippines developed an enhanced AI-powered alternative credit-scoring and risk model under its ongoing digital transformation efforts to achieve just that.

Another example (in the Philippines) involved the cement industry. One firm used AI and data science to improve the efficiency of the process of measuring the ‘compressive strength’ of cement—a crucial step in the acceptance of the product for concrete mixes used in constructing vertical structures. Using a predictive model with data transformation, modeling, and deployment, the process—which usually took 28 days, was reduced to an almost instantaneous process. Not only did this result in time savings, better resource management and operational efficiency, but it also improved the consistency of cement products.

In what is considered a traditional industry in the Philippines, this breakthrough also led to reduced annual carbon emissions, contributing to Environmental, Social and Governance goals in the process.

DigiconAsia: Given the current concerns about global oil supplies and therefore electricity costs, can AI be used to ease accessibility and supply concerns?

LG: AI can be used to help predict potential volatility to the entire power system, identify risks of failures, identify efficiency gains opportunities in generation, help us determine the best placement of power storage equipment, and facilitate the sharing of power between generators/consumers.

With today’s data science and machine learning technologies, the advanced analysis and automation of electricity generation and grid operations are becoming progressively possible. Let us also consider that access to electricity and the Internet—which is critical to sustaining and developing an economy—is still a key challenge in some parts of South-east Asia.

Therefore, the use of data and power will only increase; more data equals more energy demands. Solutions will need to be developed to maximize the use of renewables/batteries while sustaining power output from centralized sources that have economies of scale and greater energy output. AI can also be used to balance reliability, performance, efficiency and sustainability as circumstances permit.

Understanding these complexities and knowing when and how to execute tradeoffs (through actionable insights and predictions) is absolutely necessary to sustain the growth of economies and improve everyone’s quality of life.

The reality is that we can only build these intelligent decisions through the use and support of AI. Operating a power utility without it will become impossible.

DigiconAsia thanks Luis for his insights.