Haptic devices are crucial for effective human-machine interfaces (HMIs), improving interactions with robots, prosthetics, and virtual environments. Current haptic devices often use rigid actuators, limiting user dexterity. While stretchable strategies have improved conformability, the spatial resolution remains insufficient for high-acuity skin areas. Soft actuators offer an alternative, but face challenges in stretchability and actuation performance. Electrostimulation is simple but susceptible to skin impedance fluctuations. Externally force-driven actuators (pneumatic/hydraulic) are bulky. Actuators based on stimuli-responsive materials are lighter and thinner, with dielectric elastomer actuators showing promise. However, existing devices struggle to produce haptic cues compatible with skin's mechanical sensory capabilities, particularly in force and spatial perception. This research aims to develop skin-compatible haptic devices that overcome these limitations, focusing on stretchable, pressure-amplified actuators with high resolution, mimicking the skin's mechanical properties and achieving efficient closed-loop haptic communication.

The existing literature highlights the need for improved haptic interfaces. While advancements have been made in skin-integrated sensors for capturing human actions, the feedback channel from machines to users remains under-explored. Conventional HMIs heavily rely on visual and auditory feedback, neglecting the crucial role of haptic perception. Existing haptic devices often utilize rigid electromagnetic actuators, which are bulky and hinder natural dexterity. Attempts to incorporate stretchability have yielded some improvement but haven't addressed the limitations in spatial resolution and actuation performance. Soft actuators, while promising, face challenges in mechanical stretchability and actuation performance. Electrostimulation-based approaches have limitations due to fluctuating skin impedance, while externally force-driven actuators are generally bulky and challenging to miniaturize. Stimuli-responsive materials, particularly dielectric elastomers, offer a more compact alternative, but further improvements in actuation output and self-healing capabilities are necessary. Current devices fall short of producing haptic cues fully compatible with the skin's mechanical sensory capabilities, especially in force and spatial perception.

This study developed stretchable pressure-amplified electrostatic actuators with a vertical structure, improving safety by keeping the voltage-applied oil pouch away from the skin. The actuator comprises an oil-filled pouch (dielectric layers with Ag electrodes and silicone oil), a central opening covered by a stretchable silicone membrane (PDMS and Ecoflex0030 mixture), and a soft pressure-amplification structure (soft pillars). A scalable magnetic field-assisted self-assembly process was used to fabricate the pillars. The actuation principle relies on electrostatic force causing a ‘zipping’ motion of the electrodes, squeezing the oil into the stretchable region and generating pressure. The pressure is further amplified by the pillars. Thermoplastic polyurethane (TPU) was used for the pouch shells due to its ease of fabrication and textile integration. The stretchable silicone membrane's properties were optimized by varying the PDMS:Ecoflex0030 ratio and thickness. The pouch dimensions (electrode distance, height, width) were also optimized to maximize displacement, fluid conversion rate, and output pressure. Resistive pressure sensors, fabricated using porous silicone with carbon black/Ecoflex0030 composites, served as electronic skins. A user study evaluated the actuators' effectiveness in triggering tactile perception at different skin locations, comparing actuators with and without pressure-amplification structures. The study included closed-loop HMI demonstrations for surface texture and object shape recognition. The resistance changes from the pressure sensors modulated the voltage applied to the actuators to provide haptic feedback to the user. User studies evaluated the accuracy of surface texture and object shape recognition based on the transmitted haptic information.

The optimized actuators achieved an out-of-plane displacement of ~500 µm and an output pressure of ~1 MPa, exceeding skin's haptic perception thresholds. The self-assembled soft pillars significantly enhanced pressure output without increasing voltage. The actuators remained functional under 15% strain, matching the skin's elastic range. The actuator array, with a 3 mm spacing, achieved high spatial resolution suitable for all skin locations. The highly sensitive pressure sensors provided accurate data for modulating haptic feedback. User studies demonstrated high accuracy in surface texture recognition (>97.8%) and object shape recognition (100% for sphere and bar, 86.7% for cube). The actuators did not cause pain during the user studies. The fabrication processes for both actuators and sensors are scalable and compatible with textile integration.

These findings demonstrate the successful development of skin-integrated, stretchable haptic devices capable of delivering skin-compatible haptic feedback. The high spatial resolution and pressure output, combined with the stretchability, significantly advance the field of haptic interfaces. The successful closed-loop HMI demonstrations for surface texture and object shape recognition highlight the potential of these devices for various applications. The use of self-assembly for pillar fabrication improves scalability and reduces fabrication complexity. The results address the limitations of existing haptic devices by providing a solution that is both highly sensitive and compatible with the skin’s mechanical properties. This technology has potential for significant impact in robotics, prosthetics, telemedicine, and virtual/augmented reality.

This study presents pressure-amplified stretchable actuators that meet the requirements of skin-compatible haptic feedback. The high spatial resolution and pressure output, combined with stretchability and ease of fabrication, provide a significant advance in the field. The successful transmission of surface texture and object shape information demonstrates the potential for advanced human-machine interaction. Future research could explore integrating more sophisticated sensory elements and expanding the range of haptic cues to further enhance human-machine interaction.

The current user studies involved a limited number of participants. While the actuators did not cause pain in the study, further investigation is needed to assess long-term comfort and potential skin irritation from prolonged use. The object shape recognition accuracy could be improved with a larger sensor array and more sophisticated algorithms. The current design focuses on pressure feedback; future work should explore adding other tactile sensations, such as vibration and temperature.