Road congestion in India is a massive problem that is steadily getting worse. This calls for an urgent need to estimate congestion and traffic patterns on urban roads. Although much work has already been done by developed nations, but solutions for Indian cities need specific attention as the nature of traffic in India is fundamentally different from that of developed nations. This project envisages the use of mobile phones to estimate congestion and traffic patterns on urban roads. Based on the congestion metrics thus obtained, the project is developing algorithms and tools for traffic planning and management, using mobile phone as a service platform. The proposed solution strategy consists of two distinct focus areas. The first focus area deals with the problem of estimating mobile phone densities to measure prevailing congestion and traffic patterns. The second focus area involves developing algorithms for traffic routing, control and prediction, based on the estimated congestion. This work has enormous potential for applications such as dynamic route planning, peak hour rush control, routing of emergency vehicles to and from disaster affected areas, evacuation planning and traffic prediction. In addition, this work shall shed new conceptual insights into the general problem of controlling complex networks by bringing together ideas from several technical disciplines. The project also plans do a techno-economic evaluation of the system so as to understand and appreciate the extent of avoidable social costs caused by congestion.
MENTORS:
COLLABORATORS:
Remarks: Introduced measure-theoretic version of the course
Links/URL: http://nptel.ac.in/courses/108106083Remarks: Prof. Carolina Osorio (MIT), 2016
Title: Ambulance services in India: Novel models for network design and operationsRemarks: Prof. Lavanya Marla (UIUC), 2016
Remarks: 2017 - 2019
1. | A Fair Decentralized Traffic Signal Control with Good Throughput Characteristics |
2. | String and robust stability of connected vehicle systems with delayed feedback |
3. | Taxi Demand-Supply Forecasting: Impact of Spatial Partitioning on the Performance of Neural Networks |
4. | Forecasting Supply in Voronoi Regions for App-Based Taxi Hailing Services |
5. | Taxi dispatches using supply forecasting: A time-series based approach |
6. | The modified optimal velocity model: stability analyses and design guidelines |
7. | Collaborative learning of stochastic bandits over a social network |