A recent challenge is that IoT devices tend to be configured in an LLN environment. LLN has thousands of nodes with limited processing power, memory, and energy, and these nodes are typically connected by lossy links. The RPL for LLN is designed for use in resource-constrained devices in industry, homes, and cities. The primary purpose of the RPL is to provide IPv6 connectivity for many battery-powered embedded devices using low-power radios. The RPL is expected to be very active in smart factories in Industry 4.0. However, smart factories are an environment with high interference and mobility. Although RPL is designed for LLN, energy consumption may increase due to frequent parent changes in unstable networks. Adaptive RPL (ADP-RPL) calculates the degradation of link quality as the rank value increases and the parent"s change time and then reflects it in the rank value sent to neighboring nodes. In order to prevent frequent parent changes, we used a threshold that is proportional to the rank of neighboring nodes. In this paper, we evaluated the performance of ADP-RPL using a Cooja simulator. As a result, ADP-RPL shows better performance with more nodes. In order to evaluate performance in the actual environment, a test was performed in an environment in which GET and POST messages were transmitted using the CoAP protocol. RE-Mote device was used to evaluate the indoor space where interference occurred. In the real environment, the rate of ADP-RPL transmission failure and the number of parent changes were reduced compared to RPL.