Operation of modern air transportation system in a condition of congested airspace requires continuous monitoring of flight safety in real-time. One of the main components of flight safety is the risk of mid-air collision which depends on number of airspace users, structure of air traffic management, and performance of equipment used for ensuring air traffic control. In this paper, we consider estimation of risk of mid-air collision for considered airspace based on air traffic data obtained by Automatic Dependent Surveillance-Broadcast (ADS-B) technology. Two basic models of risk estimation have been used for airplanes on intersection and parallel tracks. Airspace has been represented as a set of hexagonal indexes H3. Also, H3 has been used to identify closed pairs of airspace users in a horizontal plane, which has been selected to estimate risk of mid-air collision. A univariate generalized error distribution is used as a model of airplane deviations from cleared trajectory, which is required to estimate risk of a mid-air collision. In numerical demonstration air traffic data obtained by ADS-B in real time is used to estimate risk of mid-air collision for particular airspace.