In the era of advanced technology and smart environments, automating the tasks in daily life for smart lives is under high desire. Among the tasks to be automated, electric energy meter reading and theft detection is one of the basic services. Many researchers of the field developed different frameworks of automation for the above mentioned tasks depending on different infrastructure contexts. However, the frameworks are either Ethernet /Wi-Fi or Smart phone dependent or unidirectional which are not compatible to conventional digital meters and not as per the requirements of the utility provider and the customers. This thesis summarizes a design science research conducted to design an automation framework of Ethiopian electric utility meter reading and theft detection not directly dependent on smart phones and Wi-Fi/Ethernet. To this end, this research conducted an assessment of the existing system and defined major business requirement of Ethiopian electric utility. Based on the requirements, a framework is designed for automating electric energy meter reading and theft detection. Design science research methodology is used to design and develop the prototype. The designed framework is IoT enabled automation framework using cellular networks to communicate data through SIM900A GSM module and Arduino Mega 2560 microcontroller to control the whole processes. An attempt was made to simulate and demonstrate the designed framework using Serial Monitor of Arduino IDE so as to evaluate the efficacy of the framework against the objectives defined and the requirements specified. As a result, the designed framework is an effective prototype to automate and control digital electric meter reading and theft detection. The designed automation framework of meter reading and theft detection is effective for digital electric meters. It is evaluated against defined requirements, research questions and objectives and defined solution objectives. , the framework designed is 81.8% effective and 90.9% efficient. On average, the framework has 86.35% overall performance and 86.67% theft detection performance (detection accuracy). Accordingly, the framework is well effective and well efficient, some update/upgrade for 5G networks and other meter types being recommended.