Embedded systems with microprocessors require many hardware information such as clock speed, port name, and device address to be entered into the operating system cores. In the classic compilation process for operating system kernel, all features, that the embedded system supports, need to be included in the source code. The kernel of the operating system must be recompiled for each software development process which considers related information about these features. In systems that have hundreds of different features, this process is time consuming and causes high cost as expected. Use of Device Tree data structure for the specification of physical components in embedded systems may facilitate this software development process since the kernel of the operating system is compiled just once in case of applying a development process based on this data structure. In this study, a methodology for the development of embedded software with using the Device Tree structure is introduced. New methodology is compared with existing software development approach and a preliminary evaluation is performed with taking into account pros and cons of applying this new methodology.