Fluid Flow-based Design of a System Housing for Electronic Nose Post-Harvest Application

Tomato classification is still performed manually especially by low-income farmers. This process takes a significant amount of time that may lead to worker fatigue, low precision, inconsistency, and high costs. Given these issues, an automated classification system that utilizes a camera and electronic nose-based systems may aid the local industry due to the ability of both systems to perform accurate classifications, but in order to create such a system that can be used industrially, a suitable housing must be created for it.  With that, this study provided the design process of the system housing for an electronic nose and computer vision-based tomato classification system. The design entails a housing that has four separated sub-chambers, each fitted with their respective camera and e-nose system and are capable of fitting one native Philippine tomato and simultaneously gather data for both systems for tomato classification. To efficiently make use of the e-nose systems, a fan is optimally placed to know the air flow with a 3.5cm and 6cm tomato placed along the airflow path; it was done with the aid of ANSYS, specifically CFD. By finding the optimal placement of the fan, the VOCs from the tomatoes may be properly conveyed to the e-nose system. As for the camera systems, it was optimally placed at the ceiling of the sub-chamber to properly capture the majority of the tomatoes while the e-nose system compensates for the limitations of the camera system.