Chennai’s transport authority has launched a study to map travel demand around all 41 operational Phase-I Metro stations, using anonymised mobile and GPS data to plan better first- and last-mile feeder services for a network carrying nearly 3.5 lakh passengers daily.
The Chennai Unified Metropolitan Transport Authority (CUMTA) has invited bids for a two-month consultancy to prepare big data-based origin-destination matrices for the catchment areas of the stations. The study will examine movement within a five-kilometre road-distance influence area around each station and divide the area into traffic analysis zones of approximately 0.5 sq km each.
The exercise will cover both existing Metro users and travel demand that the current network may not be capturing. CUMTA special officer I Jeyakumar said the authority needed to understand how people move between locations even when they do not use the Metro, including cases where the system serves both ends of a journey.
“Understanding such movements would require anonymised mobility data beyond Metro ticketing records,” Jeyakumar said, according to the report.
The appointed consultant will be required to procure anonymised GPS traces from smartphone applications or mobile-network data covering at least three months. The tender specifies representative coverage of around 10 to 15 per cent of the target population.
CUMTA said the data supplied to it would contain aggregated travel movements rather than individual movement histories. Machine learning-based trip-chaining algorithms will be used to convert location pings into trips and dwell periods. These movements will then be classified by origin, destination, peak or off-peak period and trip length.
The consultant will prepare separate origin-destination matrices for average daily travel, a three-hour morning peak, a three-hour evening peak and a six-hour off-peak period. The results are intended to show how travel demand changes across time and around different stations.
The study will also include a web-based visualisation platform. It will display the study area, traffic analysis zones and origin-destination movements through heatmaps and desire-line diagrams. These tools are expected to give CUMTA a station-level view of where passengers and other travellers begin and end their journeys, and how they move through each station’s surrounding area.
The data is expected to support planning for feeder services that connect Metro stations with nearby residential areas, employment locations and other travel destinations. The tender, however, does not specify the number or type of feeder services that may eventually be introduced. Those decisions will follow the findings of the study and the authority’s subsequent planning process.
The use of mobility data also places privacy safeguards at the centre of the exercise. CUMTA said individual-level data would not be supplied to the authority and that the information it receives would be anonymised and aggregated. “The data will be under our control,” Jeyakumar said.
The consultant will have to hand over the raw and final datasets, maps, geographic information system shapefiles and working files to CUMTA. The authority’s tender therefore covers not only the preparation of travel-demand matrices but also the transfer of the underlying planning material and digital tools needed for future transport analysis.
