Skin allergy detection is critical to detect allergies that trigger serious reactions such as anaphylaxis, so people can avoid allergens and reduce the risk of complications such as anaphylactic shock. Therefore, early allergy detection screening is essential to determine the risk of allergies. This research aims to develop a system to detect skin allergies caused by food, through sensors applied to human skin using the Convolutional Neural Network (CNN) model. The research steps include literature studies, data acquisition, preprocessing, learning processes, and testing. The developed system uses a camera to capture allergic reactions on the skin. Data acquisition consists of two types of data, namely primary data and secondary data. Primary data acquisition is done by taking images of normal and allergic patient skin. Meanwhile, secondary data acquisition is obtained from Kaggle. The captured images are processed by image processing and analyzed using the CNN model. The image dataset consists of four classes, namely atopic, angioedema, normal skin, and urticaria. The CNN model consists of several layers, including convolutional layers, pooling, and fully connected layers. The results of the research showed that the prototype product can detect changes in the skin surface due to allergic reactions, such as redness or swelling, quickly and accurately. Testing the learning process with the CNN model resulted in an accuracy rate of 92%. Meanwhile, the accuracy results of testing prototype products on patients with skin allergies were 93%. It shows that the system can detect types of allergies on the skin accurately and efficiently. This system provides a practical and fast solution for the public to detect allergies, while contributing to the advancement of medical technology.Keywords - social robots, adaptive learning, reinforcement learning, human-robot interaction, sensor fusion, educational robotics