Abstract
The present paper aims to analyze the characteristics of motorcycle emission levels in Makassar City, Indonesia. Some motorcycle characteristics such vehicle age, engine size, traveled distance, and brands are explored. The study uses data of the motorcycle emission test results from I/M (Inspection and Maintenance) program in the city during 2010-2012. In analyzing the influence of the characteristics on the motorcycle emission, we develop the probability failure of the motorcycle emission test using binomial logit model. Regarding the model result, the study continues to develop a relationship model between the significant characteristics and the motorcycle emission levels using regression models. The results show that the motorcycle age is the significant variable in the logit model. Further, the polynomial regression model orde-3 shows that the CO and HC emission levels increase in following the motorcycle age increasing. The results provide an expectation in developing motorcycle emission prediction model in further studies.
