This paper describes the design of a semi-automated heating and scanning system and analytic method for potato
characterization. Potatoes are heated using lamps in a heating chamber and then transferred on a movable fixture to
an imaging chamber. A non-linear model was designed to predict which potatoes have excessive sugar defects and
the model was evaluated with good results. Results from this research will benefit potato growers and
manufacturers/producers of potato-based products such as chips and fries.