 Good morning everyone, I am Yuli Depra from the University of Pisa, and today I will present an evaluation of rotation gestures in rotary versus motionless knobs. Ice-free interactions with knobs controllers, called also optic rotations, are widely used in contexts such as driving or while using professional appliances. The common mechanical knobs are very cheap and widely adopted, but their lifetime is often reduced by mechanical wear, seepage of liquids and dirt, or even vandalism. On the other hand, the virtualization of knobs on touch-screen asks for visual attention and gives less precise control. In this paper, we present the motionless knob, a fixed cylinder protruding from the user interface capable to encode the rotation gesture. The motionless knob design puts together many benefits of standard and virtual knobs, which also poses several questions in terms of technical requirements and perception action issues, such as the affordance with the user interface or the possible differences in terms of accuracy and precision between gestures and rotation. To answer those questions, we built a device consisting of a cylinder that can be set free to rotate, called rotation configuration, or locked, called gesture configuration. In the cylinder, an ISP camera is used to track the individual position of the fingertips during rotation and gestures. A further standard mechanical knob, called routine configuration, was used to validate the experimental device. Participants were asked to perform multiple optic rotations in routine rotation and gesture configuration, covering 45 and 90 degrees in both the clockwise and counter-wise direction. In its trial, we measured the mean and the variable error, the finger deviation, that is, the individual movement performed by its fingertip, and the initial position of the hand. Results on accuracy and precision show a general overshoot of about 10-15 degrees, that remains stable across the three configurations, the angle amplitude, and the direction. Sending the finger deviation gestures were found to have significantly higher values than rotations, as each finger can slide independently on the surface. The direction, instead, played a significant role regarding the initial position of the hand. In the subjective evaluation, the gesture configuration scored worse than the others in all the criteria. However, although the participants believe they perform worse in gesture configuration, objective measurements show equal mean sign and variable error. To conclude, optic rotation on standard and motionless knobs have comparable performance in terms of accuracy and precision. However, the variability of individual finger poses several issues in terms of detection and modeling, especially if gestures are not constrained. Thanks for your attention.