The University of Danang - Danang University of Science and Technology (UD-DUT) is one of the leading universites in technical training in Central of Vietnam. There are 8 lecture halls with 200 classrooms and the total area is 240,900 m2. However, it is difficult for students when they would like to find the shortest to their classrooms, cafeteria or self-study places. In order to solve this challenges, we have designed a self-study management system allowing students to find a self-study location with available space through an application on smartphone. Students can use the GPS function on their smartphone to locate their position, then choose the place they want to come, the app will show the shortest path and guide the way. This application can remind students schedule and time to return borrowed books to the library. The most important point is the application can run on both Android and iOS so that everyone can use our system. Figure 1 describes a summarized block diagram of the system. Each self-study place is monitored in realtime by a node consisting of an embedded computer Raspberry Pi 3 connected to a camera. The frame then will be processed using an background subtraction algorithm [1-3] which is embbeded in the node’s computer. Information from each node is then fowared to a gateway using LoRA wireless communication. The gateway then sends data to the cloud server by MQTT protocol. A smartphone application is developed to connect to the server and get data, then illustrate it in the app [4]. There is a menu with options to monitor self-study places, use a Figure 1: System block diagram service map or book return reminder. The application is optimized to run in as many smartphones as possible with small lag as possible. Experimental results show that when the internet is not stable, there will have a considerable delay in the application and users will have false information. Further work should solve this problem and to ensure the stability of the whole system. Feedback from users would help us to improve the user interface of the application. Deep learning algorithm will be applied to increase the accuracy of human detection at faster speed.