Automatic Risk Detection System for Farmer’s Health Monitoring Based on Behavior of Pesticide Use

Abstract

The Indonesian farmers usually use the agrochemical technology (i.e. pesticide) to prevent or control not only pests but also weeds. Unfortunately, this technique is often applied without adhering to the Occupational Health and Safety (OHS) standards, so that the risks to the pesticide exposure cannot be avoided. Nowadays, the stakeholder has the difficulties for detecting the risks. The current detection system is still performed manually by involving the related experts to assess the farmers’ behavior of pesticide use. This process often spends more times, so that it is impacted on late awareness of the pesticide exposure risk. Due to this, this paper proposed the automatic early detection (ArDeFarm) system with applying Certainty Factor method. This system uses the certainty values acquired from the related experts in order to generate the percentage of the pesticide exposure risk based on the farmers’ behavior in the pesticide mixing, loading and application. It is tested by using farmers’ behavior data collected from 100 orange farmers in Karo District. Based on the experiment, it is able to identify the pesticide exposure risk faced on farmers in the form of the risk percentage automatically, without relying on the experts. Besides, the result shows that it has been appropriate to the expert assessment.