We propose a contactless automated ordering system utilizing Kinect v2 sensing. The system applies continuous-wave indirect time-of-flight (CW-iToF) technology to detect infrared phase shifts. This sensing method generates precise 3D data by extracting 25 skeletal feature points. The system selects six key joints to formulate four three-dimensional (3D) angular features for gesture recognition. Our adaptive geometric model utilizes triangulation to calibrate interaction regions using tester height and standing distance. This sensor-driven approach achieves a recognition success rate of 96.5% at 1.5 m and 95.1% at 2.5 m. The system identifies the selected meal and instructs a robotic arm for food preparation. This architecture establishes a fully contactless and hygienic automated dining framework.