Dr ying is important for producing long shelf life food products such as vegetables, fruits and liquid foods. A batch hot air tray dr yer is widely used for food dr ying as the equipment is simple and easy to use. The problem of the food dr ying process by such tray dr yers is that it is difficult to determine the optimum drying program such as air temperature and humidity. A trial and error approach is commonly employed for determining the dr ying program, which is labor intensive and time-consuming. Another approach is to use mechanistic or hybrid models along with various monitoring devices, some of which can work as soft sensors. We have developed a laboratory drying setup, equipped with non-invasive temperature sensors, humidity sensors, a digital camera and an electronic balance. The weight of the sample, the air temperature, the air humidity, the sample temperature and the sample image can be recorded on-line by PC. A proto-type tray dryer having above-mentioned devices was also developed. As for the model, a water diffusion equation along with a heat balance equation considering the mass transfer through the interface was used. It is practically dif ficult to determine the optimum dr ying program (temperature step change) experimentally because a single drying experiment takes at least a few hours as high air temperatures cannot be used in order to avoid the quality loss of the food product. It is therefore best to use the model simulation for determining the optimum conditions with the data obtained by the dr yer with various sensing devices. We have demonstrated that the dryer developed in this study was able to monitor the color change of the model food samples continuously, which might be related to the loss of free amino acid concentrations during the late stage of drying.