Medical Internet of Things technology can effectively enable remote physiological data collection, monitoring, and transmission by integrating flexible sensors, wireless communication, and the human body in a wearable and embedded manner. Herein, a stretchable elastomer optical fiber has been developed and sandwiched with two performance of polymethyl methacrylate optical fibers to form a fully flexible polymer optical fiber sensor. The optical fiber integrated with the system mainly consists of a microcomputer, a light-emitting diode (LED) driver, a LED light source, a photodiode, and a Bluetooth module. One intelligent wearable photonic sensing system has been developed based on the Beer-Lambert law of the stretchable elastomer optical fiber for remote healthcare monitoring. Benefiting from the use of elastomer polymer materials, the sensing system features a maximum strain of more than 250%, a high-tensile strain of up to 100%, and a durability of >500 tests. Also, based on the advantage of elastomer optical fiber, the sensing part can be flexibly pasted on the skin surface as a wearable device for real-time monitoring of multiple physiological parameters. In this study, we successfully realized the monitoring of breathing pattern, heart rate, pulse, facial micro-activity, and joint activity, and the recognition of articulatory activity and knee joint activity using a 1-D convolutional neural network, with an accuracy of more than 90% for each activity recognition. Such merits demonstrate its potential as a medical toolkit and indicate promise for remote healthcare monitoring.