EMA eyes GenAI and predictive analytics to strengthen Singapore’s energy resilience

By Si Ying Thian

Proactive planning and effective integration of diverse data streams are key to ensure secure and resilient energy supply, says Singapore Energy Market Authority (EMA)’s Chief Data Officer, Chua Shen Hwee.

Energy Market Authority (EMA)’s Chief Data Officer, Chua Shen Hwee, shares with GovInsider that the biggest opportunities for EMA are to leverage analytics and GenAI to strengthen Singapore's energy resilience. Image: Chua Shen Hwee; EMA

Singapore’s energy resilience was put to proof during the Covid-19 pandemic. 

 
As EMA's CDO, Chua oversees the management and utilisation of data within EMA, including developing data strategies, ensuring data governance, analysing data for insights, collaborating with various departments, and more. Image: Chua Shen Hwee

To ensure the safety of on-site power plant operators while maintaining essential electricity generation services, the Energy Market Authority (EMA) relied on real-time consumption data from smart meters, as well as simulations to forecast short-term electricity needs for essential services.  

 

Talking to GovInsider, Energy Market Authority (EMA)’s Chief Data Officer, Chua Shen Hwee, says these real-time analytics were key for contingency planning, “to ensure that adequate and stable supply was prioritised for critical facilities such as hospitals.”  

 

The crisis management during this period highlighted the vital role of proactive planning and effective integration of diverse data streams in maintaining energy security - a lesson that now informs the EMA’s data-driven approach to energy management in resource-constrained Singapore. 

 

To subscribe to the GovInsider bulletin click here.

GenAI and analytics for energy resilience 

 

Chua believes some of the biggest data opportunities for EMA are to leverage analytics and generative artificial intelligence (GenAI) to strengthen Singapore’s energy resilience. 

 

While predictive analytics leverage AI and machine learning (ML) to generate forecasting outcomes, prescriptive analytics takes the extra step to recommend a course of action for planning and policymaking. 

 

She recognises GenAI’s “untapped potential” to synthesise and analyse both unstructured and structured data to provide deeper insights on trends, patterns and relationships - beyond just numbers. 

 

Unstructured data refers to non-numeral data, including text documents, social media sentiments, and multimedia content. 

 

To capitalise on these opportunities, EMA is investing in big data platforms and upskilling initiatives in advanced modelling and GenAI for its officers.

 

The agency will also work closely with other government agencies and industry partners to share best practices and enhance digital capabilities. 

Prioritising data integration and interoperability 

 

Chua uses the term “rubbish in, rubbish out” to emphasise the importance of data quality and accuracy when it comes to leveraging tech for energy resilience. 

 

“This task has become more complex with the proliferation of distributed data sources and the increasing number of players in the industry, such as independent retailers, electric vehicle charging providers, demand response aggregators, and solar and energy storage system providers,” she acknowledges. 

 

To ensure high-quality data and make reliable decisions, EMA is prioritising data interoperability and integration across diverse datasets, says Chua. 

 
The AI industry is considered very energy intensive, requiring significant amounts of electricity to operate and contributing to a substantial environmental footprint. Image: Canva

A holistic data view enables EMA to anticipate disruptions and optimise energy management, especially with the rise in demand from energy-intensive sectors coupled with unpredictable supply shifts from extreme weather events, she adds. 

 

The data sources that EMA is looking to integrate range from renewables and weather, consumer consumption patterns, electric vehicle charging trends, to electricity and fuel prices. 

 

“The energy system is multi-dimensional and interconnected, with various components such as generation, distribution, and consumption interacting in complex ways.  

 

“These interactions are shaped by technological advancements, regulatory policies, market dynamics, and environmental considerations,” she explains. 

 

EMA currently utilises weather data from Meteorological Service Singapore, population data from Department of Statistics, economic data from the Ministry of Trade and Industry (MTI) and other agencies to forecast electricity demand. 

 

To harness data for energy infrastructure planning, EMA also collaborates with agencies such as the Urban Redevelopment Authority (URA) and JTC Corporation to integrate energy infrastructure needs into national land use plans.  

Measuring data-driven initiatives 

 

All the data and technology are only as good as its measurable outcomes. Chua notes that she distills the effectiveness of data-driven initiatives into three key criteria: Impact on organisation's productivity, future-proof ability, and user buy-in. 

The SP app allows consumers to check meter readings by day, month or year to have an idea of their consumption patterns. Image: SP Group
 

She calls for data initiatives or systems to be “designed like Lego blocks” which are interchangeable and expandable, rather than a “single unit block” that cannot be adapted or scaled.   

 

This is especially important amidst the changing needs and requirements of users and organisations today. 

 

Finally, it is necessary to consider a user-first principle when designing these initiatives as the “final testament” of its effectiveness lies in user adoption and prevalence, she adds.  

 

Beyond EMA’s own use of data, the agency increasingly seeks to make data accessible to empower consumers and businesses to engage in energy-saving practices. 

 

For example, EMA uses smart meters that enable consumers to monitor their electricity usage real-time on the SP mobile app, encouraging them to become more efficient in their energy consumption. 

 

Additionally, the demand response programme by EMA incentivises businesses to voluntarily reduce electricity consumption during peak periods by sharing the cost savings from lower wholesale electricity prices. 

WOG collaborations as enabler 

 

Chua emphasises the value of centralised whole-of-government (WOG) platforms and interagency working groups in enabling data sharing and discovery across the government. 

 

The Singapore Government Tech Stack (SGTS), for example, enables easy exchange of data between agencies, which aggregates data collection for richer insights for policymaking and operations. 

 

As data becomes ever more important today, ensuring data security is now "mission critical” for EMA, she observes. 

 

The agency is working closely with GovTech and the Cyber Security Agency of Singapore (CSA) to ensure that EMA’s systems are strengthened with up-to-date security measures and best practices. 

 

“This helps us stay informed about new threats and vulnerabilities, and will allow us to implement timely and effective countermeasures.” 

 

EMA has also stepped up its data management and cybersecurity awareness training to stay ahead of emerging cyberthreats.