One way to do the commutation process at BLDCM is the six-step method. And hardware implementation for this method can be done in various ways. This paper will try to run a BLDCM commutation process using machine learning. Machine learning is an artificial intelligence method that can respond based on the learning process that was carried out previously through existing data sets. In this study, the training process is done offline. Then the “weight” obtained from the training is used in the Machine Learning structure which is implemented on the F28069M microcontroller. From the hardware implementation, it is proven that Machine Learning can be used in the BLDC motor commutation process.