Development of a Predictive System for Taxi Service

Taxi services have become an important part of life for most of modern society. Less than fifteen years ago, taxi services existed in a completely different form. The main competitive feature of the companies providing passenger transportation services was the literacy of drivers, their awareness of traffic in a particular city, knowledge of the shortest routes and the ability to navigate. Today, the taxi business model has taken on a completely different look. In the modern world, there are many large aggregators and platforms that provide navigation services, GPS, routes, travel time, approximate cost. Thanks to these sites, the threshold for entering the profession of a taxi driver has been significantly lowered. From now on, it is enough to have a driver's license and some driving experience, which allows you to access the taxi platform and the ability to earn money by transporting passengers. Despite the huge number of opportunities provided by the aforementioned aggregators, an important detail remains that creates a tangible disparity in the number of orders from drivers, in particular - knowledge of "hot spots" and "hot hours" in the city. In the modern profession of a taxi driver, there is a significant problem of finding and knowing the points of high demand at a particular point in time. Taxi drivers who live and work outside of large cities, in regions where there are time gaps when it is quite difficult to get an order due to weak demand from customers (for example, or during a period of time outside rush hours), face such difficulties. Large taxi aggregators do not provide their drivers with the ability to analyze orders and predict high demand at a certain point in the city at a certain time. It should be noted that in most services there is still the ability to display high demand at the moment, as well as the ability to navigate to a place of high demand. However, this functionality does not allow drivers to forecast and plan their work schedule based on this information. For this reason, there is a strong disparity between drivers who are well versed in the target city and drivers, for example, visitors from other regions.