In this study, we present a system that manages multiple unmanned aerial vehicles (UAVs) for a search, pickup, and drop mission in the 2017 Mohamed Bin Zayed International Robotics Challenge (MBZIRC). Three UAVs picked up and dropped 23 circular and rectangular targets into a designated drop box. To control the operation of three UAVs flying over an arena of 90 × 60 m, we designed and integrated a set of technologies into our system: airspace allocation, communication framework among UAVs, anticollision based on geofencing, and a token‐based prioritization for coordination. The proposed UAV system uses a single GPS and its error of a few meters is solved by means of the following component technologies: (a) flight path generator based on one reference point, (b) vision‐based redefinition of a reference point for GPS correction, and (c) calibration of flight path to update the reference point. The pickup‐and‐drop mission is conducted via color‐ and shape‐based vision processing and a magnetic gripper to pickup and drop‐off the targets. Our proposed system is able to successfully manage three UAVs, recognize targets on the ground, and drop the targets into a drop box in the drop zone. Finally, we achieved fourth place among 18 teams in Challenge 3.