These circumstances also introduce what context might be capable of during an interaction process. Desktop environments are known for being less prone to suffer from context conditions (obviously certain
Transferring the utilization of robots from the repetitive and limited tasks of the industrial environment to more complex operations for interacting with human beings has recently raised growing interest in both the research and applied technology fields. In this context, great improvements are required, not only for in-hand manipulation and exploration tasks, but also for safe operations and interactions with humans. Humanoid robots, unlike the industrial ones, are required to achieve their goals interacting with humans and their tools, adapting to the changes in the environment thanks to an autonomous learning process.

In order to satisfy these requirements, robots need to be able to perform advanced human-like manipulation tasks such as rotation, translation and in-hand grasping [1�C3].To operate in changing environments, humanoid robots need to sense and elaborate the information about the surrounding environment, while interacting with real world objects. By analyzing the force and the position at all points of contact, robots can obtain information about the weight, the stiffness and the surface of a tool and elaborate a way to complete the assigned tasks. In order to satisfy these requirements, there is increased interest in the robotic community in the development of large area or whole-body tactile sensing structures.

Without a high throughput tactile sensing system, humanoid projects strongly limit their interaction and cognitive capabilities [4]. Tactile sensing is also essential for fine manipulation tasks in humans. When our mechanoreceptors are anesthetized, like when our hands are chilled from cold weather, this results in a loss of sensing and our movements become inaccurate and clumsy. Simple operations like lacing up shoes or simply maintaining a stable grasp on an object can become very complex tasks. In order to reproduce human tactile sensing performances for fabricating sensor devices to be implemented in robot hands and bodies, several researchers have defined the guidelines and requirements Entinostat which a robot tactile system has to satisfy for performing the basic in-hand manipulation tasks.

These requirements, presented in Table 1, were determined by analyzing the human sense of touch, but even if they are almost exhaustive, they could be modified depending on the specific application in which the device would be used [3�C7]. Moreover, even if some criteria are strict and technologically challenging, a possible solution to fulfill them could be complex systems integrating different devices instead of using a single tactile sensor.Table 1.Specific requirements for the design of tactile sensor devices to be implemented on human robots.