This paper describes a concept of making interactive human state recognition systems based on smart sensor design. The token measures on proper ADC signal processing had significantly lowered the interference level. A more reliable way of measuring human skin temperature was offered by using Maxim DS18B20 digital thermometers. They introduced a more sensible response to temperature changes compared to previously used analog LM35 thermometers. An adaptive HR measuring algorithm was introduced to suppress incorrect ECG signal readings caused by human muscular activities. User friendly interactive interface for touch sensitive GLCD screen was developed to present real time physiological data readings both in numerals and graphics. User was granted an ability to dynamically customize data processing methods according to his needs. Specific procedures were developed to simplify physiological state recording for further analysis. The introduced physiological data sampling and preprocessing platform was optimized to be compatible with “ATmega Oscilloscope” PC data collecting and visualizing software.