SBIR Phase II: A Multimodal Sensor Platform for Automated Detection and Classification of Pest Insects

The broader impact/commercial potential of this project is significant. The proposed technology could drastically improve the overall effectiveness of pest management programs in various agricultural industries. The proposed system not only eliminates one of the most laborious and dreaded activities of manually inspecting insect traps, but also provides unprecedented access to accurate, real-time insect population information to make more effective pest management decisions. This leads to reduced, spatially-restricted pesticide applications, better understanding of insect pest behaviors, and enhanced biological control. The automated trap?s ability to monitor more than one species of insect not only achieves a higher level of efficiency in pest monitoring but also multiplies the technology?s value to the end users. The potential market for the proposed technology is quite broad. In fact, the technology can benefit any industry that requires regular monitoring of insect populations. Furthermore, the proposed technology could be used for various state and federal pest monitoring programs, such as the Slow the Spread (STS) Project administered by USDA to monitor gypsy moths. This Small Business Innovation Research (SBIR) Phase 2 project proposes to develop a multi-modal sensor platform for robust detection and classification of multiple insect pest species for automated monitoring of insect populations in production agriculture. The main goal of this project is two-fold: First is to demonstrate that bio-impedance sensor alone provides useful data to classify multiple insect species. A linear support vector machine classifier using mel-frequency cepstral coefficients extracted from bio-impedance data will be implemented on an embedded platform for detection and classification of two insect pest species. These bio-impedance based electronic traps will undergo large-scale field trials and be prepared for full commercialization. The second objective is to develop a multi-modal electronic trap with ultrasound, infrared and bio-impedance sensors that can simultaneously monitor four or more insect pest species. Measurement signals generated by the sensors will be analyzed to determine a set distinct features that can be computed on an embedded platform for real-time processing. These features will then be used in multi-modal sensor fusion algorithms for robust detection and classification. Different sensor fusion strategies will be investigated and the performance of each fusion algorithm will be evaluated both in controlled and field conditions.