This proposal present a Smart Training Glove that uses sensors, such as, flex sensors and pressure sensitive rubber fabric to detect hand movements and exercise load to support physical training of recovering people or regular people that require some kind of training record. The record of the details of each training session can be used to feedback on the training profile and preparation and optimize results even when the trainer cannot always be present to accompany the trainee. The structure required to assemble this device are mainly off-the-shelf components, such as, micro-controllers, sensors and communication modules, but the interpretation of the resulting signals and the optimal array of sensors is still subject to study. In the future, my intention is to optimize the use of IoT, putting together other devices and communication to a cognitive cloud to make this system capable of online feedback to the user and/or trainer to check whether the exercise objective (including posture and desired pace) is being met or a correction is required.