A smart device for detecting, monitoring the drywood termite, Cryptoterms brevis, and predicting its population density in wood-infested samples

Termites are a significant pest in many regions of the world; they attack cellulose-based materials in buildings, trees, and crops. The most significant economic losses occur to timber in structures, and a great deal of effort and money are spent to prevent damage to homes and public buildings. Termites may attack wood anywhere in a building, from below soil to the highest point on the roof. Detection of termites is often challenging due to the cryptic nature of termites, the complexity of the structure, the location of damage or termites in the structure, and available techniques. Several methods have been utilized to detect and monitor the presence of termites in buildings, from simple visual searches to technology-based or technology-assisted approaches that vary in their invasiveness and destructiveness. Termite-detection devices are tools used to locate and monitor termite activity. Several articles discuss different technologies and systems, including microwave detection, acoustic and temperature sensing, and wireless monitoring stations. These devices help professionals and homeowners to detect termite infestations early in order to implement effective control measures. Termites produce a range of sounds, usually inaudible to the human ear, through their foraging activities and communication. The use of sensitive audio equipment's could therefore be an option for termite detection in wood, where their activity might be missed. By analyzing the sounds of termites, it seems clear that they operate in the audio frequency range. The audible audio frequency range is from 20 Hz to 20 kHz. This information had inspired the authors to design and implement the present device. This paper presents a design of a smart termite-detection device that integrates sound analysis with artificial intelligence (AI) for real-time and non-invasive monitoring. The system utilities a high-sensitivity microelectronics mechanical systems (MEMS) microphone connected to a Raspberry Pi 4 to capture and analyze environmental sounds. AI models are trained to recognize the unique acoustic patterns produced by termites, particularly during feeding and movement. An innovative aspect of this system is its ability to process and refine termite-related frequencies, as well as recognizing resonant warning signals emitted by army ants, natural termite predators. Utilizing modern software supported by artificial intelligence to analyze the recorded sounds in order to detect the termite early infestations in wood; meanwhile, field trials (i.e., household furniture) were conducted in several termite- infested locations to automatically detect them. The present device is portable, easy to use, requires no wired power supply, and its components are readily available in the local market. It also offers quick response times and rapid recovery.