Abstract :
Shows how dynamic sensing, using very simple sensors, can give details on the nature of an object that at present can only be found using complex and expensive systems. The author studies the relative sensing abilities of robotic sensors, ranging from simple, but robust and reliable micro-switches, to the advanced tactile arrays. These sensors are then analyzed in terms of their data gathering capacity, and using this information it is possible to make a comparison of the capabilities of `static´ and `dynamic´ sensing systems. Based on these predictions the performance of a `dynamic´ sensor is tested and analyzed, and it is shown that outputs from this system are of a high enough standard that object textures can be easily differentiated providing in addition a rudimentary object recognition ability
