Case Study: Land Transport Authority
Singapore LTA embarks on an EDW project to meet expanding transit needs.
For densely populated Singapore, the transportation system is critical to keeping nearly 5 million citizens and socio-economic development moving. The country’s Land Transport Authority (LTA) oversees this task through traffic management and regulation of both private and public transit. The agency plans, develops and manages for short- and long-term needs to provide an efficient, people-centered system that includes roads, rail, buses, taxis and private vehicles.
Rapidly growing data volumes, however, posed a considerable challenge to LTA’s technology infrastructure. The authority captured more than 12 million records on public transport each day—and performed advanced analytics to ensure smoother travel for all. Teradata Magazine spoke with Rosina Howe, LTA’s Group Director of Innovation and Infocomm Technology, about how her organization implemented one of the nation’s largest government data warehouses to help optimize journey times and conduct strategic planning.
What are some of the unique challenges Singapore faces in transportation management and planning?
By 2020, we expect the travel demand to increase from the current 8.9 million journeys to 14.3 million per day. Given the land constraints of our small island state, our fast-paced development and a growing population, the projected demand could put a tremendous strain on our public transportation system if not planned for properly. The government launched an ambitious program, the Land Transport Master Plan, in 2008, including investing US$50 billion to further develop the transportation network over the next decade to serve the needs of the people.
How did existing technology limit LTA’s ability to carry out its mission?
Our business IT systems were designed for quick response time with the priority given to transactional processing. Since these systems are not designed to store or process historical data longer than three months, the amount of data available online was insufficient for meaningful trend analysis, long-term policy planning and data mining. At the same time, the increased sophistication and analytical complexity of more than 100 business scenarios made it impossible to meet required turnaround time. Executing a typical query accessing over 100 million records would take more than 20 hours.
What steps did you take to overcome these challenges?
In 2008, we initiated a proof-of-concept study with Teradata over 15 months to identify a suitable enterprise data warehouse [EDW] technology. Several test scenarios were run to assess the performance, scalability and workload management. The results showed a 99 percent improvement in query response times over the legacy system. Following these encouraging results, in 2009, we embarked on a full-scale implementation of our Planning for Land Transport Network (PLANET) project utilizing technology from Teradata.
What is PLANET’s role?
PLANET focuses on business data analytics for planning and statistical analysis, regulatory and operational reporting, and quality of service for public transportation. It consolidates anonymized data from disparate transactional systems, performs high-volume data crunching and presents the information from different perspectives to serve multiple purposes.
How does it benefit LTA?
PLANET is an integral part of government policy planning to support an overall vision of Making Public Transport a Choice Mode, Optimizing Road Network and Meeting Diverse Needs of the People. Today, PLANET provides LTA with advanced capabilities to perform large-scale business analytics, transport modeling for strategic planning and predictive simulation in the deployment of our various Land Transport Master Plan initiatives. Its strategic nature lies in the ability to think a generation ahead as we seek sustainable solutions to solve rising complexities around urban planning and development.
“Data-driven policy and planning decisions with fast turnaround are regarded as the critical success of PLANET.”
Group Director of Innovation and Infocomm Technology, LTA
What impact is PLANET having on policy decisions?
We have comprehensive information readily available at our fingertips for the analysis of historical performance and general traveling patterns. LTA can be more effective in carrying out regulatory audits, operational monitoring and review of existing policies.
All of this allows us to implement new measures to better address the needs of commuters. For example, we track traffic flow and behavior, passenger loading, route running times and transfer volumes for advanced trend analysis. This enables planners to determine levels of efficiency and resource optimization.
From a technical standpoint, what benefits have been realized?
The EDW offers flexibility in its architecture, enabling near-linear scalability in performance and capacity. In the processing of daily data transformation and loading, PLANET achieved a 67 percent improvement over previous methods. Based on our new system design, business data queries with at least 100 million records improved 99 percent from 20 hours to 15 minutes. New capabilities have also emerged. Within three months of implementation, 70 new analytical reports were generated for policy review, post-analysis of projects and trend patterns to optimize resource planning.
Does the system provide monetary value as well?
Definitely. We estimate we will save 19 percent annually by avoiding the cost of enhancing and maintaining various legacy data warehouses. We have also increased work force productivity by 13 percent annually by cutting previously labor-intensive tasks, such as data extraction, consolidation and validation.
Of all of the benefits realized, which elements stand out?
Data-driven policy and planning decisions with fast turnaround are regarded as the critical success factors of PLANET. Today, our expertise, knowledge and experience are complemented by data-driven analytics on the different domains of land transportation. This enables LTA to achieve excellence in our endeavor to have a people-centered land transport system.