Finger on the Pulse: Maximising Real-Time Data for Safer and Smarter Urban Mobility
This edition of Finger on the Pulse features insights from the Growing Infrastructure Course on ‘Sustainable Transport’, discussing how real-time data can be strategically used to shape safer, smarter and more resilient urban mobility systems across Asia.
Traffic congestion and transit delays exact a heavy toll on urban centres — reducing productivity, increasing pollution and diminishing quality of life for residents. To address these growing challenges, cities across Asia are finding ways to operate existing transportation systems more efficiently and intelligently.
At the 'Real-Time Data Analytics for Smart Public Transport' panel during our Growing Infrastructure Course in April 2025, industry leaders shared their experiences in leveraging real-time data to enhance urban mobility through smarter transportation systems. The session brought together Dr. Chin Kian Keong (Senior Consultant, ST Engineering), Dr. Renrong Wang (Head of Data Science, Grab) and Andy Chiang (Managing Director, STRIDES). Professor Archan Misra, Vice Provost (Research) of the Singapore Management University (SMU), moderated the discussion.
From the insights shared, here are three approaches that cities and transport operators can adopt to unlock the full value from real-time data.
Photo: (L to R) Dr. Chin Kian Keong, Dr. Renrong Wang and Mr. Andy Chiang
1. Start with the Purpose, not Technology
Effective use of real-time data begins by defining the why: what problems are we solving, and who benefits from the insights? As Dr. Chin highlighted, real-time data must ultimately serve the foundational goals of all transport system: safety, efficiency, accessibility, equitability, inclusiveness and sustainability.
In fact, over-engineering data can lead to "noise" rather than clarity. For example, collecting data every five seconds may be unnecessary, and even counterproductive, if the system does not require an immediate response. Instead, it is important to identify the appropriate level of responsiveness and align data latency, frequency and granularity accordingly.
"There is a fallacy that the more real-time [a piece of data is] the better. Because I've often found that after a certain point, you actually get noise in the data," shared Prof. Misra. Andy reinforced this point, sharing that STRIDES’ need for real-time data tends to be highly selective and risk-based: "[It always sounds good to have real-time data], but it can be costly. Hence, you must be very clear why you need the real-time data."
Determining the type of data that needs to be real-time depends on the operational needs. Andy provided an illustrative example of how his team adopts a risk-based assessment to address this question, focusing real-time monitoring on systems where immediate consequences are most severe. For systems that degrade gradually over time, daily or weekly monitoring might be sufficient.
Similarly, Dr. Wang shared how dynamic pricing strategies benefit from more granular passenger demand data to enable accurate fare adjustments. To enhance fare predictions, the system also incorporates real-time weather data from third-party sources. This weather data helps identify when demand spikes are due to weather conditions (like sudden rain causing more ride requests) rather than genuine market demand, allowing the system to filter out these weather-induced anomalies and produce more accurate and reliable fare predictions.
As Dr. Chin aptly adds, the key is ultimately ensuring the data collected is accurate and timely for its purpose.
2. Translate Data into Actionable Insights
The real challenge often lies not in collecting data, but in transforming that raw information into actionable and explainable insights that improves decision-making. For the Singapore Mass Rapid Transit (SMRT) Corporation, which operates Singapore's oldest metro lines, this means distilling complex system information into metrics that different stakeholders can immediately apply to their work. "We provide data that is processed to give them information they want," Andy explained. "For example, if they are interested in [financials], we talk about [financials]. If they are interested in reliability, we talk about reliability.”
This targeted approach recognises that different team members have different priorities. Financial controllers need cost metrics, maintenance teams require performance indicators and operations staff depend on reliability measures; all derived from the same underlying data but presented differently to facilitate action rather than cause analysis paralysis.
Real-time data analytics can also be used predictively. Dr. Wang explained, "We use both explicit and implicit signals at Grab, like user feedback or skipped pick-up point, to anticipate demand and optimise how our drivers are matched with passengers." Forecasting across different timeframes, from the next 10 minutes to weeks in advance, also helps match services more effectively to user demand.
Furthermore, predictive maintenance is one of the most valuable applications of this data translation process, particularly for ageing infrastructure. Rather than following generic manufacturer maintenance schedules that may not be customised for Singapore’s high-heat, high-humidity environment, operators can develop customised approaches based on actual performance data. This data-driven approach becomes particularly critical when managing Singapore's oldest transit infrastructure. The solution involves instrumenting these systems, tracking performance metrics and developing data models that evolve with the equipment's age and condition.
3. Build Trust Through Responsible Data Sharing
To maximise the impact of real-time data, cities must foster ecosystems where public and private players can collaborate on shared insights. But data sharing needs to be underpinned by clear governance and purpose. For example, Singapore's Land Transport Authority mandated open APIs aspects such as bus arrival and occupancy data, enabling third-party apps to improve commuter experience while preserving data integrity.
Groups like the Community of Metros (CoMET) offer secure, anonymised benchmarking across transit operators, helping to identify gaps and inform long-term strategy. Andy further emphasised their principle of purpose-driven data sharing: "We share data for a purpose. To improve the industry and to uplift reliability [of our transportation networks] ... We try not to share data that can cause alarm." This selective approach ensures that shared information contributes to system improvements and industry-wide learning rather than causing undue or misplaced public concern or revealing competitively sensitive information.
Another important external data source highlighted by the panellists was weather data. For Grab, weather significantly impacts both rider demand patterns and driver availability. Dr. Wang explained, during heavy rainfall, its platform observes a substantial increase in booking requests as people avoid walking or using other weather-exposed transit options. The situation is further exacerbated by conflicting driver behavior, as while some drivers remain active during these periods to capitalise on higher demand, many others stop driving to avoid the hazardous conditions, thus reducing supply. For SMRT Corporation, weather data helps with both operations and infrastructure planning. This is especially crucial because many transit systems were not designed with Singapore’s climate conditions in mind.
Beyond improving current operations, responsible data sharing enables a new class of "proactive" or "anticipatory" mobility services. Prof Misra shared a case study by his colleagues, who analysed smart card transactions generated by automated card fare systems during commuters’ journey. They found empirical evidence that people do not always take the shorter route and instead, prefer longer routes to avoid crowding or minimise transfers between metro lines.
By modelling these nuanced preferences and estimating travel times segment-by-segment, researchers can reduce prediction errors and generate more accurate journey estimates across different route options. These insights enable transport authorities to move beyond using mobility data merely for policy planning and instead create adaptive systems that anticipate and respond to commuter behaviours in real time.
Photo: Professor Archan Misra
The Road Ahead
As Prof Misra noted, transportation is not just about moving people efficiently; it is also about delivering economic and social benefits that can transform communities. For Asia's diverse urban landscapes, real-time data represents a vital component that can maximise these benefits while minimising infrastructure costs.
One significant challenge along this journey remains the integration of systems. Modern transit networks comprise components from numerous manufacturers, each with their own data formats and monitoring systems. A single metro line includes rolling stock, trains, signalling systems, platform screen doors, track systems and power systems — all produced by different Original Equipment Manufacturers (OEMs). Despite this complexity, these components must function seamlessly to ensure safety and efficiency. Creating unified analytics across these fragmented systems is therefore both a technical challenge and a strategic imperative for the future of urban mobility.
As Asia's cities grow and become more complex, intelligent use of real-time data will be essential to ensuring their mobility systems remain resilient, dynamic and prepared for future demands.
At Infrastructure Asia, we identify regional infrastructural needs and connect them to best fit solutions suited to Asia's diverse urban environments. We work with Singapore-based solution providers offering a comprehensive range of smart mobility technologies and urban traffic management systems, to support the evolving requirements of transport authorities and public transport operators across the region.
