Singapore’s EVe harnesses the power of data with help from NTT DATA, AWS

In alignment with Singapore’s ambitious sustainability objectives, the Land Transport Authority (LTA) of Singapore is intensifying its efforts to spur the adoption of electric vehicles. This contributes to the nation’s goal of achieving 100 percent cleaner energy vehicles by 2040, and the effort leverages Amazon Web Services (AWS).

LTA has set up EV-Electric Charging Pte Ltd (EVe) to manage the deployment of up to 12,000 electric vehicle (EV) charging points distributed across 2,000 Housing Development Board (HDB) carparks. EVe will coordinate the upgrading of supporting electrical infrastructure for EV charging.

The scope of EVe’s responsibilities extends beyond this initial deployment, as EVe is also looking at evaluating and expanding the EV charging infrastructure as part of its business growth. This includes scaling up the number of EV chargers and extending their reach to encompass not only HDB carparks but also other publicly accessible carparks. In essence, EVe plays a vital role in orchestrating the comprehensive integration of EV charging infrastructure, and is contributing significantly to the realization of Singapore’s sustainable and forward-looking transportation landscape.

EVe’s transformation journey

Since its inception, EVe recognized the pivotal role of data and has become a data-driven organization. The initial step involved calling up a comprehensive tender to establish a secure, scalable, and flexible data platform. The platform allows EVe to make well-informed decisions, optimize resources, and strategically plan for future growth.

EVe’s immediate goal is to harness the power of data to monitor the use and operational metrics of EVe charging stations spread across Singapore. Through this initiative, EVe aims to seamlessly integrate data-driven insights into its resource planning, ensuring efficiency and the realization of its strategic goals.

EVe encountered challenges in consolidating near real-time data from diverse and disparate sources, such as charge point operators and government data platforms. The intricacies extended to storing this huge volume of data and, more critically, deriving actionable insights from it. Seeking an innovative solution, EVe initiated a meticulous tendering process, inviting vendors to propose software as a service (SaaS)-based solutions that could effectively address these challenges.

After careful evaluation, EVe selected NTT DATA‘s e-Mobility Data Platform, built on the robust infrastructure of AWS, as the ideal choice to tackle EVe’s complex data integration and analytics needs. Using the Open Charge Point Interface (OCPI) standards, this platform adeptly ingests data from various sources. Once assimilated, the data undergoes a secure and sophisticated lifecycle, such as storage, analysis, and visualization, seamlessly executed through AWS native services.

Architecting the solutions

NTT DATA’s e-Mobility Data Platform uses key AWS services (shown in Figure 1) such as Amazon Simple Storage Service (Amazon S3) for storage, Amazon Athena for querying, Amazon EMR for distributed data processing, Amazon Redshift for data warehousing, and Amazon Managed Grafana and Amazon QuickSight for intuitive and comprehensive data visualization. This cohesive integration of cutting-edge technologies makes sure that EVe not only overcomes its initial challenges but also propels itself to prepare for the future.

Figure 1. NTT Data e-Mobility data platform.

The detailed solution architecture on AWS is shown in the following Figure 2:

Figure 2. EVe’s solution architecture.

1. Ingestion: The ingestion architecture employs multiple techniques to securely accept user data, both batch and real time, thereby providing flexibility. A network firewall provides a secure endpoint for AWS Transfer Family to accept user files over SFTP into a dedicated Amazon S3 raw landing bucket. Additionally, an Amazon Elastic Container Service (Amazon ECS) Fargate application behind an NAT gateway collects real-time data by calling external APIs, with the response data also landing in Amazon S3 raw storage. AWS Database Migration Service (AWS DMS) can also sync changes from a user relational database, outputting data files to the same Amazon S3 repository. Having a unified raw landing zone allows the same validation and transformation process to handle these diverse ingestion sources. Event notifications on new Amazon S3 objects trigger downstream ETL processing by AWS Step Functions orchestrating EMR Spark jobs to prepare landed data for analysis. Supporting batch upload, real-time APIs, and database sync provides flexibility on ingress before standardizing data for processing.
2. Processing: Step Functions orchestrates EMR Spark jobs to transform the raw landing zone data into processed stage data. Amazon S3 event notifications trigger Step Functions executions whenever new data lands. EMR Spark jobs validate and convert the raw data into formats optimized for analytics such as Apache Parquet. The processed output data lands in an encrypted Amazon S3 bucket that serves as the stage area of the lake. This stage data layer contains intermediate, consumption-ready data with table metadata stored in the AWS Glue Catalog. The raw upload data is retained in the landing zone with versioning enabled to maintain history.
3. Analytics: The EMR Spark jobs orchestrated by Step Functions triggers Amazon Redshift to efficiently load the Apache Parquet stage data using the Copy command. This parallel data transfer uses the Amazon Redshift MPP architecture for fast analytics data loading. The aggregated data lands in Amazon Redshift tables optimized for business intelligence workloads. Amazon Redshift serves as a performant data warehouse enabling easy SQL-based analysis through QuickSight dashboards, Amazon Managed Grafana visualizations, and ad-hoc queries with Athena. This analytics layer provides consumption-ready aggregate data tailored to the use cases driving the architecture. Retaining raw and stage layers allows reprocessing as needs evolve over time using the orchestrated EMR Spark jobs.
4. Consumption: QuickSight and Amazon Managed Grafana provide interactive dashboards and reports to visualize the analytics data, delivering actionable business insights. These visualizations are embedded into a web application powered by Amazon ECS Fargate containers fronted by an Application Load Balancer (ALB). Amazon CloudFront secured by AWS Shield and AWS WAF accelerates delivery of the containerized application. Amazon Route 53 routes end users to the CloudFront distribution. Additional security measures include AWS Secrets Manager for housing credentials and AWS Certificate Manager for SSL certificates. After identity verification through AWS IAM Identity Center, users can uncover insights from these fully managed visualization services while the backend auto-scales to demand.

Conclusion

NTT DATA strategically selected AWS as its cloud partner, driven by the extensive breadth, scalability, and flexibility of services that AWS provides. This strategic alignment empowers NTT DATA to seamlessly expand its SaaS offerings, leveraging the comprehensive portfolio of services AWS offers. The partnership not only makes sure of the effortless expansion of capabilities but also positions NTT DATA to adapt and innovate efficiently in the evolving landscape.

Under the SaaS model introduced by NTT DATA, the conventional Capital Expenditure (CAPEX) is replaced with Operational Expenditure (OPEX). This shift enhances cash flow dynamics and contributes to a notable reduction in the total cost of ownership for EVe. The e-Mobility Data Platform solution, a result of this collaborative effort, plays a pivotal role in optimizing EVe’s operations by significantly reducing the time to insights, cutting costs, and facilitating the swift integration of new data sources and features.

NTT DATA and AWS are committed to help customers harness the value of data with their technical expertise to realize their business outcomes and pave the path towards a sustainable future.