Posi-Feel(tm) is our patent pending method of giving the user a sense of feel and haptic feedback. Traditionally robots that convey a sense of force to the user use a "force feedback" system, in which the force the robotic gripper is applying is communicated to the controller, which exerts the opposite force on the user. Posi-Feel(tm) turns this method on it's head using "position feedback force control". Instead of the robot communicating force to the controller, the controller measures the users input force and communicates this force to the robot. The robots grip force is then controlled by the users input force, regardless of the grippers position. The users sense of feel is achieved because the robot is feeding back it's gripper position to the controller. For example, if the robot grabs a soft object with a large force it will squish, and then un-squish when the force is removed. Using a Posi-Feel(tm) controller the user can control this grip force and feel the "squishiness" of the object.
Posi-Feel(tm) is so seamless that most users don't even know it's active until they try a robot without it. The feedback generated is so intuitive that manipulations are possible that couldn't be accomplished without such feedback. Position feedback allows the user to grab objects of various sizes and irregular shapes and allows the user to know when the object might be slipping our of his or her grip, while force control allows the user to grab incredibly fragile objects without consciously adjusting their grip.
The use of force control instead of position control has huge implications for programming. Unlike servo based robot grippers, Posi-Feel(tm) allows the robot to grab objects without knowing their size, or alternatively allows the robot to grab object of various sizes using the same program. The user determines the force needed to lift an object, either by inputting the force directly into the program or recording the force in manual mode, and the robot will grab the object using that force regardless of size. This allows for decision making on the part of the robot. The robot can sort by size, for example, or determine if the object has been successfully grabbed (see this in action in experiment "didIGetIt"). This approach can be used for anything from heavy wooden blocks to empty egg shells.
Taken as a whole, Posi-Feel(tm) control allows for simplified, flexible programming. The user can easily implement Artificial Intelligence into their robot in ways not possible with servo based robots. The possibilities for a robot with this technology are exponentially greater than position based gripper control.
Posi-Feel(tm) works by using two separate control systems, one for the robot and one for the controller. Two force sensitive resistors (FSR) on the controller gripper measure the force applied by the user. A microcontroller applies adjustments per a pre-defined control system and sends a torque command to the robot's gripper.
The position of the robot gripper is measured by a potentiometer. This value is compared to a potentiometer on the controller gripper, and a microcontroller applies a control system to send a torque to the controller motor. This torque is designed to hold the controller gripper in exactly the same position as the robot, thus achieving a sense of feel.