Ride sharing is becoming a growing business. It took Uber five years before reaching 1 billion rides. But, six months later Uber, hit 2 billion rides and the company is continuing to grow. In fact, Uber has partnered with McDonald’s to provide delivery service to consumers in major Florida cities. In addition, General Motors has created a plan to rent cars to Uber drivers for only $179 per week. Now, MIT researchers have developed a new algorithm that could disrupt the entire taxi industry using Uber’s ride-sharing technology.
Researchers discovered that 3,000 four passenger cars could replace 98 percent of all New York City taxis. In fact, researchers suggests ride-sharing apps like Uber and Lyft have the potential of eliminating large-scale traffic, pollution and energy usage in New York City.
Research led by MIT Professor Daniela Rus of MIT’s Computer Science and Artificial Intelligence Laboratory also found that the average wait-time would be less than 3 minutes.
Future of ride-sharing
Rus explained in a statement that ride-sharing carpooling would help transport multiple people at a time resulting in fewer trips, and ultimately, less money.
“A system like this could allow drivers to work shorter shifts, while also creating less traffic, cleaner air and shorter, less stressful commutes.”
Currently in New York City, there are approximately 14,000 taxis. But, nearly all of the city’s transportation demands could be replaced by only 2,000 ten-person vehicles, according to the algorithm.
Researchers were able to make this discovery by examining 3 million taxi rides. Researchers then rerouted the cars based on incoming requests. In addition, researchers had inactive cars moving to areas with high demands.
With that being said, researchers argue that ride-sharing programs are unable to handle the algorithm’s predictions. At the moment, ride-sharing applications need to be designed so that multiple requests can be sent at one time before an algorithm can design an efficient route. Consequently, that makes the real-world application of their algorithm difficult, according to Rus.
“A key challenge was to develop a real-time solution that considers the thousands of vehicles and requests at once,” says Rus. “We can do this in our method because that first step enables us to understand and abstract the road network at a fine level of detail.”