To create an immersive experience in extended reality (XR), providing effective and versatile touch feedback is essential but remains a technical challenge. Specifically, imitating everyday touch interactions, such as grasping and pressing, poses considerable challenges and requires highly interdisciplinary research approaches. Current solutions in research and industry rely on devices that are bulky, burdensome, tethered, limited in haptic expressiveness, and costly, degrading the system's wearability and resulting in a less-immersive experience.

Our work introduces a design strategy for compact, flexible, and salient haptic feedback centered on a 30-μm-thick inflatable chamber that naturally conforms to the fingertip []. We leverage the benefits of soft electrohydraulic actuators, including high speed, broadband, energy efficiency, and untethered operation, while using soft flexible materials that are widely available and inexpensive. To minimize fluidic losses and enable high bandwidth, a soft electrohydraulic pump mounted close to the fingertip actuates the chamber through a mechanically transparent fluidic channel. We demonstrate an exemplary feedback system with a wide actuation bandwidth from steady state to 500 Hz, well-matched to the human sense of touch. It can generate tunable hydraulic forces easily perceivable at the fingertip, with a peak force of 8 N and a steady-state force of 3 N. These capabilities enable the proposed actuation paradigm to deliver tactile sensations mechanically comparable to real touch experiences. Our design can enrich various applications in XR, including medical training, online shopping, and social interactions.

Additionally, we propose a novel method to assess the quality of fingertip haptic feedback by employing a commercially available artificial finger that closely emulates the form factor and sensing capabilities of real human fingers. Attaching the actuation system directly to this artificial fingertip enables a direct comparison between the haptic sensations it can create and those generated during typical hand interactions. This assessment method can standardize evaluation protocols for haptic devices, enabling rigorous comparisons between actuation paradigms.