Uber Go Scheduling at Bangalore Airport — A Simulation using Go

If you live in Bangalore as I do, it is highly probable that you must have been to the Bangalore airport at least once and it is highly likely that you have used their Uber Go service at the airport arrivals.

I recently returned from a family vacation to Go a, and we used the Uber Go service upon our arrival to head back home. I noticed that they have an interesting scheduling and queuing system in place as this service is the most popular among travellers.

The Problem

The dynamics at the Uber Go pick up station is briefly as follows.

People queue and wait there for Uber Go taxis to arrive. It appears that they allow pickup by up to 3 taxis concurrently. The Uber Go drivers pick a fare, drop them to their destination, pocket their fare and usually circle back to the airport for more pickups. There is a larger queue for waiting drivers and I believe, a kind of token system.

We can assume the incoming taxi traffic rate is fairly constant — though in reality it may actually vary depending on the time of the day.

As a queuing and concurrency problem, this is interesting as it is a system of producers and consumers. (We can think of either the Taxis or the Passengers as producers/consumers — it does not matter).