Urban Air Mobility (UAM) utilizes electric vertical takeoff and landing (eVTOL) technology for low-altitude urban travel, easing ground traffic congestion. However, its complex flight environment demands higher pilot situational awareness (SA). This study designs a human-machine interface (HMI) based on eye-tracking technology to enhance task performance and attention to critical information. A simulated flight experiment compared two conditions (HMI vs. no HMI) across autonomous flight: distraction, attention refocusing, and emergency response tasks. Results show that the HMI reduced reaction time (22.6% in emergencies), increased fixation duration (21.4%), improved attention allocation (18.1%), and decreased saccades (27.3%). These findings suggest that an optimized HMI effectively directs attention, reduces cognitive load, and improves emergency response. This study provides design insights for UAM HMIs and highlights the role of multimodal interaction (visual, auditory, haptic) in enhancing flight safety and usability. Future work could integrate AI and AR to improve adaptability.