Resolving battery status and customer matching to create 24/7 drones based advertisement system

We consider the problem of devising an advertising system that uses drones that carry advertising signs. The system should satisfy the following constraints:•There are several types of customers the system should identify and classify the current customers.•Each drone can carry a one/double-sided advertisement sign suitable to only one or two types of customers.•The system must use ultra-light drowns whose battery is limited to few flight minutes after which a hovering drone must land for recharge.•The system must use the minimal number of drones that are needed to maintain one drone hovering 24/7.This leads to an algorithmic problem wherein the system must decide A) when to send the hovering drone to recharge and B) which of the recharging drones should replace it. This should be done such that: 1) a maximal number of customers see a sign matching to their type, and 2) maintaining the battery power for the 24/7 mode. The proposed solution is based on a way to classify the discharging curve of the hovering-drone and the charging curve of the currently charging drones. Based on these battery statuses and the current two largest sets of customers, the system maximizes the number of customers that are matched with a suitable sign yet maintains the battery power suffices for the 24/7 hovering mode. The results, obtained by a system of three ultralight drones in a supermarket, show that this algorithm works with high efficiency. We describe how the system was implemented and how the results were obtained.