Currently, the Internet of Things (IoT) and Artificial Intelligence (AI) are influential technology fields that have a wide impact and hold an important position in the future. Contributing to the wide network of physical devices linked together, the long-range (LoRa) network control system has become very popular and widely applied in many fields. With low power consumption and transmission distance between 2 nodes up to 5km, LoRa is very suitable in IoT applications, smart home, or reading temperature, humidity, water level values, etc. in areas with difficult terrain and low energy consumption. However, the LoRa hardware architecture is only for point-to-point connectivity and does not support mesh multipoint connections. For this reason, this article presents hardware and algorithm development for LoRa that supports mesh architecture. Through the microprocessor, routing algorithms are embedded in the hardware to route packets through multiple intermediate nodes from the source node to the destination node. Data packet will automatically find the shortest path through the continuously updated routing table. Thereby, data is sent from the sensors to the Internet Gateway in the fastest way. Compared with the LoRa network, besides the advantage of extending the wireless transmission distance, mesh networking also helps to limit the rate of packet loss because the packet has many paths between the source node and the destination node.