Traducing Software Voice-Pattern-Detection Applications into Hardware Systems

Most of the voice-controlled systems sold are based in an expensive software pattern detection technology. We have developed a pure hardware-based system able to detect chains of letters and sounds to control house environment with light, temperature and door locking devices among others. This application is divided in 3 parts, one for sensing and transducing, other for pattern detection and a final stage for output generation. The first one is achieved roughly, using a letret microphone as the main sensing element. The second is amodular section in which we use PLLs, and simple circuitry to detect the letters. Finally the third involves power stages and signal generation for specific applications. Our application can be integrated to software systems to increase its visual appeal and interactivity with the user. The results gave us ideas to implement this kind of hardware interfaces in museums and augmented reality systems improved with computer graphics interfaces in which even if the low cost gain is lost, novelty and interactivity are surpassed. We have presented a new improve commercial system using analog and digital electronics to produce cheaper and innovative devices.