In this paper, a Machine to Machine (M2M) Relay mechanism is proposed to save energy consumption by reducing the number of repetition in a Narrowband Internet of Things (NB-IoT) system, which is a 3rd Generation Partnership Project (3GPP) user equipment (UE) category for low-power, low-complexity and battery-powered devices. A NB-IoT system can support the access of a large number of devices ranging over a coverage area larger than the typical cell in a cellular mobile network by adopting the repetition-based transmission to increase the receiver processing gain, i.e., to increase the chance of successful data transmission. 3GPP NB-IoT RAN1 already defines that the number of repetition will binary exponentially increase when the receiving signal strength decreases. This may induce surplus repetition, i.e., unnecessary energy consumption. Moreover, for devices in the edge of a cell, the inferior radio condition will incur abundant repetitions, depleting the battery energy very quickly. Thus, by using the Machine to Machine (M2M) Relay that is already developed and defined in 3GPP standard, a novel mechanism for adequate relay UE selecting and optimal relay scheduling is proposed to effectively reduce the energy consumption but maintain the system throughput and UE QoS. The extensive simulations is executed in a setting of 1000 devices distributed according to a Poisson Point Process over a cell of the radios of 125 transmission distance units. If a UE located within a transmission distance unit to the eNB; it needs only one repetition to successfully transmit data. Simulation results show that the repetition number will drop dramatically in all simulation cases. The less number of transmitting devices, the higher saving ration of repetition. For transmitting device density 0.1, 0.2, and 0.3, the repetition saving ratio is 0.7, 0.69, and 0.65 respectively. This indicating that by the proposed M2M relay scheduling mechanism, when the transmitting device density is 0.3, the system can save a huge 65% energy consumption.