C.A.R.E. (Cognitive Ambience Respiratory Engine)

The C.A.R.E. (Cognitive Ambience Respiratory Engine) project operates as a mobile, voice-interactive robot designed to monitor and purify indoor air while navigating your living space. At the core of the system is the ESP32 Dev Module and the ESP32 S3 Box 3, which process incoming voice commands using on-device WakeNet keyword recognition combined with external cloud intelligence like the Deepgram API and Gemini AI Text-to-Speech engine. When a user speaks a command, the ESP32 interprets the instruction and coordinates physical responses across multiple subsystems. For mobility, two stepper motors are controlled via dedicated A4988 motor drivers, translating digital step and direction pulses from the microcontroller into precise wheel movements to drive the base platform around the room. Simultaneously, an onboard 12V DC fan draws air through a roll of cotton or filter material mounted to the chassis, actively filtering particulates as the robot roams. Because high-current inductive loads like motors and fans can introduce electrical noise or voltage sags that disrupt sensitive digital logic, the circuit design relies on separate power lines to keep the microcontroller and communication modules stable during continuous operation.