TCS is part of the Government Data Architecture initiative to enable secure data sharing and usage across public service, built upon an integrated data management framework that manages the data lifecycle effectively from acquisition to destruction. It is a one-stop central sensor data platform that provides end-to-end data sharing services with exploitation environment.
Before the introduction of GDA, government agencies shared data on bilateral terms. The process was inefficient and led to the following issues:
- Lack of data and sub-optimal data granularity – agencies do not have visibility on what data is available or if the required data was being collected in the first place. Furthermore, granularity of data collected or shared may not be adequate to satisfy agency’s use cases.
- Data fragmentation – each requesting agency negotiates and receives sensor data from providing agency on bilateral arrangements it introduces ambiguity on schemas and sharing modality.
- Data authoritativeness – there are no common or longitudinal comparable set of data available to agencies which potentially creates disagreement over the authenticity as each agency has their own vested interest in publication or use of the set of data.
- Governance and security – there are no trackable and transparent enforcement over the request and approval workflow, which is key to reduce unnecessary exfiltration of data.
- Lack of processing capabilities – raw data can be voluminous and complex for agencies to work on. When agencies are not equipped with technical capabilities nor have sufficient funding support, deep-dive studies or analysis on their relevant subject matters are not possible.
- Excessive turnaround time – as a result from customised bilateral arrangements, the time required to scope and fulfil the data request are excessive. It takes between 6-13 months to complete the data sharing process.
To overcome the prevalent issues faced and to align with the Government’s effort to become data-driven to the core, where public agencies can make use of data to formulate better policies, deliver more personalised and integrate services to the public and support businesses in innovation and growth, GDA was launched. It is built upon the Integrated Management Framework and lays out the organisational structures and technical infrastructure required to facilitate efficient data sharing of clean and authoritative datasets. It does so by designating and building 1) Single Sources of Truth (SSOTs) that acquire clean and maintain high quality core data, 2) Trusted Centre that fuses and distributes core data and 3) central platforms for data users to request, download and analyse datasets.
TCS platform offers a seamless data request workflow that simplifies the user experience from data discovery to exploitation; users can view and request the dataset(s) directly with the routing of approvals being tied to attested identity sources within the government directory to fulfilment of request on analytics environment where the user can work directly on their scoped dataset. The platform greatly enhances the exploitation experience and contributes to faster turnaround in policy planning/analysis/design, operational service delivery and better enforcement outcomes.
TCS envisions a two-pronged approach for the next bound. Firstly, to expand and extend usage of core data within and outside of whole of government agencies. Secondly, to enhance the platform features and interoperability with other agency platforms. The focal of the first approach is to promulgate the use of sensor core data. It is vital for designated datasets to remain relevant and useful to optimise limited resources invested to upkeep the standards of the existing datasets. Within the government agencies, TCS aims to set definitive tagging of fused data to improve data mapping and regularise processes and updates to improve data lineage and service delivery. This helps to raise requestors’ confidence in data quality and trust in its utility for use in longitudinal studies.
Externally, TCS acknowledges the growing need of data from private sector to discharge its functions and to improve its offerings to general public. Similarly, private sector also require government-held data to improve their service delivery and make informed decisions. Hence, the next step is the extension of government’s data management model to enable public-to-private data exchanges. The second approach emphasizes on the platform capabilities to increase automation and interoperability with requesting agencies’ systems for managing data quality and fulfilment of data requests. This is with anticipation that the platform adoption and load will continue to grow with finite resources.
The key change is to enhance system-based interoperability between agencies’ systems and client software (eg. Qlik/Tableau/PowerBI connectors) which involves the creation of a data fabric or data layer on a new Snowflake Data Cloud, to allow data to be shared without raw data exchange. Essentially it allows TCS and users to propose error fixes or data quality enhancements back through to source provider thereby eliminating the frequent lag between error detection and fix completion. Other works include feature enhancements to improve connectivity, access times and security.
What Makes Your Project Innovative?
The GDA is a whole of government new initiative. Before GDA was introduced, government agencies mainly work off bilateral data sharing model. The model poses many fundamental issues, and it was time and effort consuming.
What is the current status of your innovation?
The primary objective was to enable secure data sharing and usage across the public sector. TCS managed to deliver the targeted outcomes, by satisfying agency use cases for better policy planning and improved operations and service delivery through the provision of high-quality data with a fast and secured access. However, the team did not stop after meeting the objectives, it went on to explore technical or non-technical (user experience) improvements and possible data exchange gaps.
Through continuous conversations with different stakeholders and polling for feedback from data users, TCS can identify the areas to improve and work on. The platform feature and enhancements are very much driven by the needs of agencies and users. Hence, when an area is identified to be work on, it goes through the process of developing proposals and implementations again.
Collaborations & Partnerships
The data providers/collectors are imperative in the data sharing process because data acquisition is the first step of the GDA construct. Data could originate from private companies (such as SP services for utility consumption data) or government agencies (such as municipal feedback data). It is essential that the data source is accurate and comprehensive, hence TCS works very closely with the respective parties to communicate feedback on data completeness and accuracy.
Users, Stakeholders & Beneficiaries
Government agencies directly benefited from the cost/time savings from bilateral sharing and avoided for having to set up engineering team or vendor service request.
Private companies benefit from the public-private data exchanges to have quicker access to agency publications and statistics to make informed business decisions and improve the service productivity and efficiency.
Citizens given convenience as companies/agencies have already obtained the required data to attend to their needs.
Results, Outcomes & Impacts
Within 2 years of operationalisation, the platform has supported more than 760 users across 37 agencies and fulfilled more than 170 data requests within 7 working days.
As the platform load is expected to increase over time with limited resources, the plan is to roll out progressive self-help features, which will result in a flatter but continued growth to 1500 users across 60 agencies by Mar 2027 and fulfilment of 100 data requests per year within 7 working days by FY26.
Challenges and Failures
SSOT may not want to be part of GDA to take on the responsibilities to clean, verify and collect additional data to close the data gaps if it is not within the agency’s mandate. To tackle this, resourcing is provided at the ministry level for the redesigning and establishing of new infrastructure to take on the added responsibilities. The data sharing policies are documented in the Instructions Manuals to dictate source agencies to share data by default, except for legislatively locked data. This has helped to influence SSOTs onboard GDA more readily.
SSOTs are often constrained by their existing long-term contracts with vendors, so the changes on process, systems, storage, transfer modality, etc to align to GDA requirements may require additional resources. TCS works with SSOT to come up with plans to deliver the GDA targeted outcomes. The trade-offs for cost-benefits are weighed and decided before investment of resources funded centrally.
Conditions for Success
Policies must be in place and aligned to drive, support, and govern the data sharing process and practices.
Supporting governing ministry/office/working committees, in this case, Government Data Office, which oversees the GDA, helps to expand outreach faster and convince stakeholders to adopt GDA. It also helps to dissect the overall strategy into more relevant and achievable targets for TCS.
Manpower and funding resources are vital for implementation of any changes. If additional resources are not awarded to the SSOTs and TCs, then the parties will not be motivated to work on initiatives outside of their agency’s mandate.
Uers have to be equipped with the knowledge on how to use data to derive meaningful insights. Hence, data literacy of officers is an important factor that will contribute to the success of GDA. TCS contributes to this by way of creating tutorials, blogs and videos, to aid users to upskill and uplift their capabilities.
This idea of having a multi-tier scalable data architecture, putting in the necessary data infrastructure, and having the data owners being responsible for the quality of the data in order to enable impact beyond their own unit can be replicated in any organization that aspires to be data ready and driven, no matter big or small. To the best of our knowledge, the GDA idea has been adopted loosely by a few Singapore government agencies so far. However, we do note that in order to replicate this model successfully, it requires all the relevant stakeholders to collaborate closely.
The aspiration of having data driven governments often require the close collaboration of policy, technology and stakeholders, where everyone involved is required to put in the due effort to make this end-to-end venture a success. At the same time, it is important to prioritize tasks where the efforts can be best focused on in order to gain small wins rather than boiling the ocean. As such, over the course of our journey, there is a need to remain patient and focused and build the fundamentals well before going on to building more things.
- Developing Proposals - turning ideas into business cases that can be assessed and acted on
- Implementation - making the innovation happen
- Evaluation - understanding whether the innovative initiative has delivered what was needed
25 January 2023