How 3,000 Uber rides could replace New York's entire taxi fleet
New York City could operate with no more than a few thousand taxis or rides if we took carpooling more seriously, researchers have suggested.
When commuters share their rides through carpooling, the cost of individual transport lowers and there are potentially fewer cars on the road which can cause increased congestion and pollution.
Modern ride-hailing services including Uber and Lyft, in which passengers can hail a ride through mobile devices and also share their vehicle with others, have caused controversy and are often less-than-popular with traditional taxi firms, but researchers say these services could have a real impact on how many vehicles are needed on our city roads.
Currently, New York City has issued almost 14,000 official taxi medallion licenses, but according to researchers from MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL), the city could make do with no more than a quarter of rides.
The team has developed a new algorithm which crunched the numbers and, assuming carpooling becomes more popular, MIT has come to the conclusion that NYC only needs 3,000 four-person cars to serve 98 percent of transport demand in New York City.
The average waiting time would be less than three minutes for a ride.
Led by Professor Daniela Rus, the MIT CSAIL group's algorithm also suggested that if two-person carpools were in use, 3,000 could serve 94 percent of demand, and only 2,000 vehicles would be needed to cater for 95 percent of demand if they carried 10 passengers each.
The algorithm uses data from three million NYC taxi rides to work out in real-time where the hotspots are for transport requests and how to proactively reroute cars to the areas with the highest levels of demands, a feature MIT says improves service speed by up to 20 percent.
"To our knowledge, this is the first time that scientists have been able to experimentally quantify the trade-off between fleet size, capacity, waiting time, travel delay, and operational costs for a range of vehicles, from taxis to vans and shuttles," says Rus.
"What's more, the system is particularly suited to autonomous cars, since it can continuously reroute vehicles based on real-time requests," the researcher added.
Today, carpooling services are somewhat limited -- especially as requests need to be in before a route can be determined -- but the researchers say that in the future, the algorithm could be used to rematch requests to different vehicles, including those with larger capacities, while also keeping in mind passenger cost, time and convenience.
Rus calls the system an "anytime optimal algorithm," which means that it improves as more data is fed into the algorithm and over time with frequent use.
"Ride-sharing services have enormous potential for positive societal impact with respect to congestion, pollution and energy consumption," says Rus. "I think it's important that we as researchers do everything we can to explore ways to make these transportation systems as efficient and reliable as possible."
The research was published in the journal Proceedings of the National Academy of the Sciences (PNAS).