This article describes the mechatronic design of a delivery octarotor drone equipped with safety features including protective bumpers for its propellers, parachute, and retractable gear. A GNSS-RTK unit provides extensive pose-refinement for trajectory tracking, while an RGB-Depth Camera provides dis- tance measurements from any present obstacles. A deep learning algorithm classifies any objects into certain categories (i.e., balloons, drones) followed by a trajectory avoidance algorithm. An onboard computer handles these sensor measurements and communicates with the autopilot for adjusting the trajectory. At the same time, it provides HD-live video feed to the base station and runs any onboard ROS-services. The resulting drone can safely land within a circular area of radius 1.5 m, while flying over, under and around obstacles or buildings. The carried payload is handled by a dual linear actuator gripper that releases it upon the drone’s precise landing.