Nairobi is one of the most heavily congested cities in Africa. Each day thousands of Kenyans make the trip into Nairobi from towns such as Kisii, Keroka, and beyond for work, business, or to visit friends and family. The journey can be long, and the final approach into the city can impact the length of the trip significantly depending on traffic. How do traffic patterns influence people’s decisions to come into the city by bus and which bus to take? Does knowing the traffic patterns in Nairobi help anticipate the demand for particular routes at particular times?

The aim of the competition is to create a predictive model using traffic data provided from Uber Movement and historic bus ticket sales data from Mobiticket to predict the number of tickets that will be sold for buses into Nairobi from cities in "up country" Kenya.

The data used to train the model will be historic hourly traffic patterns in Nairobi and historic ticket purchasing data for 14 bus routes into Nairobi from October 2017 to April 2018, and includes the place or origin, the scheduled time of departure, the channel used for the purchase, the type of vehicle, the capacity of the vehicle, and the assigned seat number. Zindi competitors will be allowed to create their own customized traffic datasets using the Uber Movement platform.

This resulting model can be used by Mobiticket and bus operators to anticipate customer demand for certain rides, to manage resources and vehicles more efficiently, to offer promotions and sell other services more effectively, such as micro-insurance, or even improve customer service by being able to send alerts and other useful information to customers.

The solutions to this challenge are the first step towards solving Nairobi's traffic problems. We look forward to taking this journey with you!

This competition is sponsored by Uber , Mobiticket, and insight2impact.