Manual and automatic acoustic detection of red palm weevil hidden activity.Recurrent monitoring of naturally weevil infested palms through larval development.Optimize system sensitivity by systematic threshold of observer and machine input. The red palm weevil (RPW), Rhynchophorus ferrugineus, is a major pest of various palm species including dates and Canary palms. The weevil's larvae develop within the tree stem and crown, damage its vascular system and eventually cause the death of the tree. Early detection of the RPW infestation is particularly challenging as the pests develop within the palm, well hidden from human eye. Our work focused on the acoustic detection of RPW larvae activity. Young date and Canary palms were naturally infested by exposure to adult males and females RPW and were monitored acoustically and visually for several weeks. A piezoelectric sensor was used to capture the larvae's distinct sounds that propagate through the fibrous palm tissue. To determine whether the trees were infested, the sounds were recorded in situ and diagnosed by a human listener and by a software ("machine"). All experiments were concluded by dissecting each palm to assess its actual infestation.Human and machine detection were both efficient in detecting infested trees, with average "true positive rate" (sensitivity) of 75% (maximum 88%) and 80% (maximum 95%) for human and machine detection respectively. The sensitivity was lower during the early phase of infestation (39% and 33% respectively), and significantly improved as larvae developed.Manual and automated acoustic monitoring was found feasible for monitoring young palm trees. Manual filtering of external stimuli such as wind and ambient noise were sufficient to enable detection in an unshielded natural environment.