Xerox Innovations in Transportation
Xerox Innovations in Transportation
|Today, 64% of all travel done by urban population and the total amount of urban kilometres travelled is expected to triple by 2050. Delivering urban mobility to cope with this increasing demand will thus be one of the key themes of the future. Xerox Research Canter India is embarking on a new initiative to solve the mobility challenges of the future. They plan to facilitate this by seamlessly combining different modes of private transportation with public transit through multimodal trip planner with ride sharing, and using sophisticated data-driven analysis and efficient planning tooptimize both commuter travel time and infrastructure utilization. These solutions are aimed to cater to the large scale transportation demand that is currently unsatisfied by public transit alone. Two such innovations are briefly described below.|
Scheduling for Public Transportation
Scheduling buses is an important factor in running the operations of a public transport. The schedules of buses must be in synch with the demand for buses to ensure optimal utilization of buses and improved commuter satisfaction. The conventional pipeline of scheduling has been to first estimate the travel patterns of users using a manual survey or by other automated means (ticketing data, mobile phone data, vision based methods etc…) The data is then processed to understand the mean behaviour of the commuters. The travel patterns are then used to produce schedules in such a way that buses make more trips during peak periods and the route between popular bus stops are served more often.
In cities with large populations and limited infrastructure, the travel patterns and numbers can vary a lot between days. Hence, a schedule built on mean demand pattern becomes irrelevant for a particular day resulting in sub-optimal utilization of resources and commuter ire. In order to address this problem, Xerox research has come up with novel techniques that builds a dynamic schedule that responds to changes in the demand by introducing small alterations. Through innovative techniques of optimization, the alterations are kept small, so that the commuters are not inconvenienced. The resulting dynamic schedules translate into substantial savings for the transit agencies while keeping the commuters happy.
Urban Mobility: Multimodal Transportation with Ride sharing
Public transports have traditionally large share of total passenger travel, contributing up to three-fourth of total passenger kilometres travelled. With increased urbanization, improved affordability and rise in living standards, the transport demand is set to shift from traditional modes to new-age modes like hailing a Uber, ride-sharing, shared bikes, city metros, private shuttle services etc. A typical trip in near future will be an optimal combination of various modes to reach destination in a sustainable, cost-effective way.
Xerox Research is working towards creating such an ecosystem to provide optimal and sustainable route planning involving various operators and providing access to a large number of transit modes. The aim is to provide a service that partially fulfils the need for a certain type of travel. The end users will greatly benefit from single platform that can help them access large number of modes of transportation. Xerox Research has also come up with novel algorithms and systems for optimal and dynamic route planning for on-demand modes of transport. These can be used for both optimizing the operations from a service provider point of view as well as for improving the user experience for on-demand modes.
In particular, we focus heavily in different modes of shared transport, including ride sharing and carpooling. With increasing fuel prices and rising traffic congestion, ride sharing and carpooling are gaining in popularity and importance all over the world. The different modes of ride sharing also include shared employee transport provided by employers in service delivery organizations in emerging markets. Ride sharing brings up a number of interesting questions, such as how to route the vehicles satisfying everyone’s timing and walking/driving constraints, how to split the costs among the people sharing rides, how to ensure fairness in routing and pricing, how to learn the compatibility among users, so as to recommend the right groups to car-poolers, and how to design an efficient real-time system that can track rides efficiently for better prediction of arrival times and improved planning. We at Xerox Research are working on systems and algorithms to solve these problems, so as to facilitate ride sharing and increase its adoption by end users.
2014: Aditi Raghunathan, Ankit Kumar
2015: Adit Rustagi, Anumeha Rai, Koushik Chattopadhyay, Mohammed Feroz.
- Abhishek Tripathi, Vaibhav Rajan, Narayanan U. Edakunni, Predicting arrival times of vehicles based upon observed schedule adherence, 2015
- Narayanan U. Edakunni, Sharmistha Jat, Diagnostic tool to measure sensitivity of service quality of a transit network on weather conditions, 2014
- Narayanan U. Edakunni, Aditi Raghunathan, Dependency graph to predict demand of buses by combining historical and real time data, 2014
- Sharmistha Jat, Koyel Mukherjee, Narayanan U. Edakunni, Pallavi Manohar, Demand Responsive Cab Re-routing, 2014
- Koyel Mukherjee, Ankit Kumar, Pallavi Manohar, Narayanan U. Edakunni, Sharmistha Jat, Static and Dynamic Routing and Scheduling of Shared Transport, 2014
- Arpita Biswas, Koyel Mukherjee, Pallavi Manohar, Partha Dutta, Dynamic Pricing Systems And Related Methods, 2015.