An Energy-Efficient Ride-Sharing Algorithm Using Distributed Convex Optimization

With a rising need for transporting people from one place to another especially within urban spaces, many mobility-on-demand (MoD) systems like Uber, Lyft and the like have burgeoned to provide swift and efficient mobilty services within just a few taps through our smartphone devices. Although carpooling, an inherently fuelefficient concept, has penetrated into this market, commercial ride-sharing systems (e.g., Uber Pool, Lyft Line, etc.) do not perform the vehicle dispatch optimization in a fuel-efficient manner for the rider pick-up and drop-off phases of travel. Moreover, present ride-sharing algorithms have a centralized computing back-bone – making online ride-sharing optimization problems compute heavy, and thereby once again energy inefficient. With the rising level of tailpipe CO2 emissions, it is prudent to focus on fuel efficiency to curb climate impact from warming.