Digital Tech | Analytics and AI

The Rise of AI: Driving AI Transformation for Actionable Intelligence

Tan Shi Hui

Engineer, Data Analytics and AI, Digital Systems, ST Engineering

To keep up with the rapid digitalisation pace, organisations are racing to implement AI models and strategies which will help in transforming operations, enabling them to achieve accelerated growth. AI opens the door to more opportunities, enabling leaders to reshape their businesses as they are able to better appreciate the business dynamics, leading to higher adoption rates of AI strategies. In this interview with our data analytics and AI engineer, Tan Shi Hui, we get to understand what AI transformation entails and how it can be applied to achieve sustainable organisational growth.

Q: What does AI transformation entail and why is it important to pursue it?

AI transformation involves digitalisation of processes which drives cultural change, governance and ethics within organisations. The focus is on maintaining healthy risk postures, yet striving towards continuous improvement in results.

During challenging times, the increasing shift towards virtual and remote collaboration garners the need for AI models and strategies to be emplaced for business continuity, to minimise disruptions while ensuring business growth. With the implementation of AI, enterprises and organisations will also be able to solve complex problems with data insights, driving smart and sharp decision making.

Hence, it is important to integrate AI into business processes in order for organisations to stay ahead of competition, and introducing automation to increase the level of effectiveness in business workflows and processes.

Q: What are the 3 value propositions which AI transformation can help organisations to achieve?

AI transformation can take on multiple fronts, like helping organisations to alleviate manual processes, pre-empt and predict outcomes based on insights from data, and allocating resources efficiently.

1. Reduction of manual processes

Disintegration of data systems can result in manual processes as data points are not communicating to one another, hindering the organisation’s ability to visualise how processes can be streamlined.

With the use of natural-language processing (NLP), the ultimate goal of which is to enable computers to make sense of human languages provides many capabilities. The creation of chatbots, language translator and targeted advertising reduces the need for labours manpower to perform equivalent or more productive work that brings in more businesses opportunities.

With the use of computer vision (CV), concerned with the theory for building artificial systems that obtain information from images, such as a video sequence, views from multiple cameras, or multi-dimensional data from a medical scanner assist humans in object detection. This technology assists one to identify subjects at specific locations and send alerts back. In the healthcare, this technology is able to assist new doctors who do not have much experience to correctly identify malicious cancers as such reduces the time and effort needed to perform accurate judgments.

By connecting the data systems, AI can help in system integrity, system security and detecting non-intuitive relationships in the multi-dimensional data space, which is not easily detected by humans.

In the cyber space, AI can help detecting abnormal behaviours and malicious codes before the attack manifest throughout the systems. Such actions can heavily impact the organizations to a total halt that disrupts the businesses, amounting to huge losses. In many cases, it leads to leakage of confidential information that are misued by malicious users such as accounts being hacked and misuse of credit cards.

In addition, performing sense making in the high-dimensional space can be challenging for human hence tapping on AI to trigger alerts is a necessity especially in today big data era. AI comes in handy in fraud domain space in detecting non-intuitive relationships between entities. This provides a holistic overview of all entities before coming to a conclusion of any fraud transactions.

In turn, organisations will be able to have a full picture of what’s happening on the ground and implement necessary changes to increase business efficiency.

2. Creating meaningful data insights

Many organisations have a wealth of data but there’s a knowledge gap in understanding how to use it, thereby crippling the ability to visualise, or to gain meaningful insights from the data lake. With AI models in place, they are able to give us insights into the past, understanding the future and advise on possible outcomes which aids in more informed decision-making, achieving business objectives.

3. Efficient allocation of resources

AI helps in allocating resources efficiently, shortening time required to process workflows. For example, in control rooms, the use of AI enables public security agencies to deploy the right resources in the shortest amount of time. For hospital operations, with AI embedded, it enables optimised bed allocation and with data predictability, and it empowers hospital management to be equipped with insights that helps with decision-making especially in unexpected times of crisis. 

Q: Any best practices to follow during the AI transformation journey?

During AI transformation, it is critical to have agile and flexible practices with adequate understanding of customers’ needs so as to ensure that the recommendations can be applied and integrated across different business functions. This way, organisations will be able to attain their desired ROI/outcome by taking into consideration compliance policies and cultural shifts through an agile iterative approach.

ST Engineering Modified CRISP-DM Methodology with MLOps diagram

We improvised a model comprising of both agile delivery and CRISP-DM methodology to quickly understand the customer’s business needs and translate that into AI/ML models when required. Incorporating MLOps, pre-modelled AI/ML models will undergo continuous testing, integration and validation, coupled with constant and iterative feedback loops. Time-to-production of AI/ML models is accelerated, which enables rapid deployment, enhancing work processes.

Q: Any success stories to share on how AI has transformed business operations?

At ST Engineering, we have transformed a myriad of industries from land, sea to air using AI.

On the ground, we have implemented Electronic Management Advisory System (EMAS) in our public roads for maintenance, with innovative idea as an option such as EMASpedia – i.e., developing an AI-based Knowledge Corpus which is capable of monitoring the health of the EMAS systems to help better predict impending breakdowns within or across systems that will contribute to a Smart Intelligent Traffic System. Apart from this, we have also optimised national water network to meet National Water Demand and react to contingency. Another interesting project which we have embarked on is to provide on-demand autonomous shuttle service at Gardens By the Bay.

At seas, we have Unmanned Surface Vessels (USV), where we predict potential USV engine failures of up to 24 hours in advance and Next Gen Vessel Traffic Management System with comprehensive maritime analytics, that can predict anti-collision warning for ships in Singapore Straits and improve decision time for maritime security.

For the aviation industry, we have also developed Aircraft Engine Predictive Maintenance where prediction of CFM engine failure can be up to 60 flight cycles in advance.

Sign Up for E-News

Enjoy new insights and articles, delivered directly into your inbox

Show Buttons
Contact Us
Email Us
Hide Buttons