Food waste is a severe problem in Indonesia. Based on data from the Barilla Center for Food and Nutrition, Indonesia ranks second in food waste per person each year. The FAO data also shows that there are 13 million metric tons of food wasted, equal to the consumption of 11% of Indonesia's population (around 28 million people) each year. One factor in food waste is the food consumer's misinformation of the expired date. Not all food items have a "Best-Before Date,"especially for meat sold in traditional markets and some supermarkets. Even if it exists, it could be less accurate because it only follows the Standard Operational Procedure of the supermarket, which is food must be sold within three days. Therefore, our research aims to design a device that can detect meat expiration time using IoT (Internet of Things). This device will be developed using gas sensor MQ137 for detecting the level of NH3 released by meat. Detection results from the sensor will be sent via the internet to Firebase. The data will be processed using polynomial regression to produce a prediction of chicken meat's expiration time. Lastly, prediction time can be easily accessed through an Android application. In this paper, we present our initial works to develop the IoT device for predicting the expiration time of the meat. The paper presents the design and performance evaluation of the device.