Peatland is important to rural communities’ livelihood due to its potential for aquaculture and agriculture. Nonetheless, human activities such as slash-and-burn can greatly increase forest fire risk, which can release a great amount of greenhouse gases and carbon dioxide into the atmosphere. To sustainably manage and restore peatlands, the Internet of Things (IoT) system can incorporate with Cyber-Physical System (CPS) for peatland management. In this study, an IoT system is deployed in the peatland to monitor the ground water level (GWL) and upload it to the server for the machine learning (ML) process. The trend of GWL will be modelled, and the CPS using the developed ML model will control the peatland rewatering process. As a result, the peatland condition can be monitored in real-time, and the risk of forest fire can be mitigated through rewatering automation before the GWL drops to a critical level.