Location-based services have expanded rapidly over the years, but Global Positioning System (GPS), as the basis for outdoor positioning, has limitations that prevent it from functioning properly indoors. In situations where it is necessary to obtain greater accuracy in indoor environments, alternative solutions to a geo-positioning system must be implemented and technologies that meet these requirements must be used. In this research, the objective is to create an IoT-based system for indoor asset tracking and identification using machine learning, hence this paper presents the design and development of the electronic devices capable of communicating with each other to send information to a central system that determines the location of assets in a controlled environment, and machine learning is used as a method of location estimation. Considering that there are multiple external factors that affect the accuracy of traditional position estimation algorithms, deep learning is implemented and the data obtained from the evaluation of the model's performance in a controlled space is analyzed. It is important to highlight that in this project it was relevant to have control of the multiple variables that affect the performance of the system, for this reason the hardware, firmware of the scanning stations and TAGs using BLE and the iBeacon protocol were designed and developed.