| Proj No. | B3256-251 |
| Title | Semantic Pipeline for Human-AI ATC Simulation |
| Summary | This project seeks to develop a semantic pipeline to enhance human–AI interaction in air traffic control (ATC) training simulations. Current RSAF training depends heavily on human role-play, which is resource-intensive and often inconsistent. The proposed system will integrate speech recognition and semantic extraction to accurately parse local ATC phraseology, targeting more than 90 percent transcription accuracy and more than 95 percent extraction of key aviation terms. AI pilot agents will then generate contextual, phraseology-compliant responses to support realistic bidirectional communication. By delivering automated assessment and consistent feedback, the system will improve training efficiency, strengthen communication proficiency, and reduce reliance on manpower in ATC training operations. |
| Supervisor | Prof Cheng Tee Hiang (Loc:S1 > S1 B1C > S1 B1C 107, Ext: +65 67904534) |
| Co-Supervisor | - |
| RI Co-Supervisor | - |
| Lab | |
| Single/Group: | Single |
| Area: | Intelligent Systems and Control Engineering |
| ISP/RI/SMP/SCP?: | ISP: ME5 (Dr) Lim Yong Zhi Deputy Head of AETHER Republic of Singapore Air Force lim_yong_zhi1@defence.gov.sg |