There are nearly 13,000 taxis operating in the New York City and they could all be replaced by a sizable 3,000 ride-sharing cars under an exclusive carpooling program, according to Computer Science and Artificial Intelligence Laboratory (CSAIL) of MIT. Commuters who use the ride-sharing services instead of taxis are expected to play an important role to reduce pollution, traffic congestion as well as fuel consumption so carpooling can be an effective solution to some of the major problems of the city.
CSAIL calculated that with an average 2.7 minute waiting time, as many as 3,000 cars that can accommodate four passengers (who are travelling to the same destination) can actually reduce the demand for taxis by 98 percent. The researchers used public data published by the University of Illinois for the calculation. One of the most important aspects of this system is that, this may bring dynamic re-positioning of the vehicles in New York City based on the real-time demand. This will speed up the whole process by twenty percent.
Researchers also studied the extreme carpooling scenario where two thousand larger automobiles with 10-person capacity could meet up to 95 percent of the demand. The system could also deploy vehicles with different passenger capacities according to demands, something that will ensure maximum utilization of the cars. The extended capacity vehicles can be used for taking passengers to a concert or a big sports event, replacing the smaller four-passenger fleets. With Lyft and Uber working aggressively to expand their operations, the study comes as a positive sign to address traffic congestion and pollution for New York City through carpooling.
The system uses a detailed algorithm to assign cars based on incoming requests making sure that each vehicle is getting the best assignment to respond to the calls and taking passengers to their desired locations within the shortest period of time. Once cars are assigned, the remaining ones are directed to other areas where demand is high.
One of the major challenges for the researchers was to make the best use of real time data to determine how each vehicle could be assigned so that they could utilize their capacity to the fullest according to Professor Daniela Rus, the researcher who led the CSAIL study. But the team developed an effective system that allowed them to plan the deployment of cars according to demand while considering the road networks to ensure faster travel experience to the passengers, reducing their waiting and travelling time, she added.
Rus also said that the system would fit more effectively for autonomous cars as it reroutes vehicles according to the real time incoming requests. Lyft and Uber are already working on developing such algorithms which the companies are possibly keeping confidential as trade secrets.