With growing popularity of Internet of things (IoT) more and more sensors are getting connected to public network. Identification of sensors is a permanently growing concern. The conventional way of adding a cryptophytic module along with a sensor is vulnerable to direct attacks; unauthorized sensor replacement, malicious signal insertion, etc. are possible. A new sensing architecture inspired from nature’s sensing systems can prevent such direct attacks. The idea is to develop sensors with unique and unclonable characteristics and enroll the unique sensor characteristic into the verifier, for instance a cryptography module. As the verifier is tuned to the specific characteristic of a particular sensor, direct attacks are difficult. In order to develop such highly secure sensing modules, unique and unclonable sensors that are similar to sensors found in nature are required. Unique and unclonable sensors respond to specific mechanical parameter and each sensor output is different from others. Random structural variations present in the sensor structures are sources of the uniqueness. As random structural variation is difficult to measure, model or duplicate, such sensors are unclonable. Such sensors can be identified from their output. Design, modeling, simulation, fabrication, testing and application of such unique and unclonable sensors are discussed in this chapter.