Smart aitime vending machine
The most important intention of this thesis includes undertaking modern technology by designing and implementing a smart airtime vending machine known as a self-service airtime vendor machine that will come as an additional method apart from the current airtime selling and buying methods which are mobile money, banking, airtime agent that involves a lot of issues and risks like theft of money and airtime loading mistakes and errors. It will help rural citizens to buy airtime using a coin where a customer has to enter the mobile number using a keypad then inset coin in airtime vending machine, and automatically the machine dispenses the airtime equivalent to the amount inserted.
The proposed methodology consists of an IoT system where the customer will access the vending machine by inserting it into a coin to buy airtime. This research consists of three main parts, the first part is the interconnection of IoT hardware components that build the entire circuit and are linked to the cloud via GPRS/GSM communication technology, this part involves sensing components, data processing components, and actuators components. The second part consists of coding using Arduino IDE that makes IoT system hardware operational and the last part is data processing and analytics using python programming and regression as a machine learning technique. The system monitoring is done through wireless radio, the cloud data storage is secured and can be easily accessible by authorized users via a web interface. The battery is used for powering the system and the solar panel for recharging the battery. All transaction data are recorded and given date returns the day type between working days, weekends, and the session of the day.
The expected results of this research include an IoT system that is developed and implemented to help both airtime agents and customers to sell and buy airtime using coin-based self-service airtime vending machine and the model that analyse machine transaction data using Python programming and regression as a machine learning technique.