Speech is a major form of communication among human beings. An impairment to the speech generation components such as the brain, larynx and vocal tract diminishes its generation. The speech impairment forces the person to communicate through gestures and expressions which may not be understood well to other individuals. Here, we propose an artificial speaking system for the people with speaking difficulties. It is based on capturing the information from gestures through flex sensors and accelerometers. For every gesture, there is a specific meaning which is used for day to day communications. The message is kept in a voice playback and recorder integrated circuit. For every action, the flex sensors give signals to the microcontroller. The artificial voice is generated by pre-loaded speech through loudspeakers.