Public Transportation Demand Analysis: A Case Study of Metropolitan Lagos

TL;DR

This study analyzes Lagos's public transportation demand using ticket data, demonstrating that dynamic scheduling can significantly reduce passenger waiting times and improve urban transit efficiency.

Contribution

It introduces a data-driven demand analysis and compares fixed versus dynamic bus scheduling, providing actionable insights for urban transit planning in rapidly growing cities.

Findings

Dynamic scheduling can reduce waiting times by up to 80%

Demand patterns vary by time and location, enabling targeted scheduling

Station segmentation helps tailor bus schedules to demand characteristics

Abstract

Modelling, simulation, and forecasting offer a means of facilitating better planning and decision-making. These quantitative approaches can add value beyond traditional methods that do not rely on data and are particularly relevant for public transportation. Lagos is experiencing rapid urbanization and currently has a population of just under 15 million. Both long waiting times and uncertain travel times has driven many people to acquire their own vehicle or use alternative modes of transport. This has significantly increased the number of vehicles on the roads leading to even more traffic and greater traffic congestion. This paper investigates urban travel demand in Lagos and explores passenger dynamics in time and space. Using individual commuter trip data from tickets purchased from the Lagos State Bus Rapid Transit (BRT), the demand patterns through the hours of the day, days of the week and bus stations are analysed. This study aims to quantify demand from actual passenger trips and estimate the impact that dynamic scheduling could have on passenger waiting times. Station segmentation is provided to cluster stations by their demand characteristics in order to tailor specific bus schedules. Intra-day public transportation demand in Lagos BRT is analysed and predictions are compared. Simulations using fixed and dynamic bus scheduling demonstrate that the average waiting time could be reduced by as much as 80%. The load curves, insights and the approach developed will be useful for informing policymaking in Lagos and similar African cities facing the challenges of rapid urbanization.