Background
The advancements in tactile perception and feedback technologies have propelled the interaction between humans and the digital realm, spurring innovative applications across various fields, including virtual reality, augmented reality, disability assistance, and communication. At present, surface tactile feedback devices predominantly operate through two mechanisms: electrical stimulation and mechanical vibration. Electrical stimulation works by directly stimulating nerves with an electric current, thereby transmitting signals to the brain. Nevertheless, this approach has several drawbacks, such as inconsistent user experiences, skin discomfort, and temporary sensory desensitization. In comparison, flexible mechanical actuators convey tactile information by inducing skin deformation. This method offers greater safety and does not necessitate a close fit, thus enhancing the user's freedom of interaction with the environment. However, these actuators commonly employ flexible functional materials like dielectric elastomers and piezoelectric materials, which require kilovolt-level high-voltage power supplies for actuation. This not only poses safety hazards but also restricts the capacity to edit multi-dimensional tactile feedback information. Consequently, there is an urgent need to develop tactile feedback devices that can be driven by low voltages and possess high programmability to enable complex emotional interactions and dynamic command interactions.
Research Progress
Recently, to address this issue, the research group led by Qian Xiang from the Division of Intelligent Instruments and Equipment at Tsinghua Shenzhen International Graduate School proposed a flexible electret tactile feedback actuator based on multi-layer variable stiffness polydimethylsiloxane (PDMS) elastomers. This actuator consists of a five-layer sandwich structure (as shown in Figure 1A), including a polyethylene terephthalate (PET) encapsulation layer, an indium tin oxide (ITO) upper electrode layer, a fluorinated ethylene propylene (FEP) electret film layer, a gold electrode layer, and a PDMS stiffness regulator layer. In the group's previous work, it was found that by adjusting the parameters of a single-layer PDMS support layer, the strength and the frequency of mechanical vibrations generated by the electret actuator could be effectively controlled to encode tactile information. In this study, the group introduced the concept of variable stiffness composites by selectively changing the crosslinking density of multi-layer PDMS elastomers.
The experimental results show that the actuator has excellent output characteristics, achieving: (a) ultra-low driven voltage, capable of generating perceivable tactile feedback force at as low as 5V drive voltage; (b) high gain, with a gain of 1.06 mN/V at 200V drive voltage and 0.98 mN/V at 35V drive voltage (as shown in Figure 2E); (c) wide frequency range, with the actuator's bandwidth ranging from 50 Hz to 450 Hz (as shown in Figure 2J), covering the main frequency range sensitive to touch.
On this basis, this research designed and implemented a six-point actuator array system. This system features a unique four-dimensional (4D) tactile programming ability, encompassing time (start, duration, and stop times of vibration), position, amplitude, and frequency (as depicted in Figure 1D). Furthermore, this research presented two application scenarios of this actuator array system.
First, the "rhythm" parameters of tactile feedback, namely the regulated frequency, vibration duration, and pause time, were correlated with the intensity of emotions. The "smoothness" parameters of tactile feedback, that is, the rate of change of amplitude, were associated with the degree of emotional pleasure. As a result, three rhythm modes (R1, R2, and R3) and four smoothness programming modes (S1, S2, S3, and S4) were obtained. Through the combination of these modes, emotional information such as "Passion," "Nervous," "Sadness," and "Relaxation" can be superimposed on the dynamic output of Braille characters (as shown in Figure 3).
Secondly, by modulating the vibration timing and amplitude levels among the points of the array, an illusory tactile sensation of flow was induced in users. The key parameters for regulation include the actuation flow direction (AFD) between two actuator units, the actuation onset time (AOT), the overlapping vibration time (OVT), and the actuation amplitude (A). Based on these, this research defines five motion directions: forward/backward, left/right, clockwise, diagonally forward, and arrival (as depicted in Figure 4).
The test results indicate that, without prior training, the system attained an average accuracy rate of 64.6% in emotional interactions. After the participants underwent the learning mode, the recognition rate increased to 95.8%. In the context of navigation interactions, the average accuracy rate for direction commands was 94.2%.
Future Prospects
Future work may focus on further enhancing the resolution of emotional coordinates to elevate the complexity and subtlety of haptic simulation. To assess real-world applicability, especially for visually impaired users, we are actively pursuing collaborations with organizations serving the blind community and planning targeted user studies. Additionally, the multi-stiffness soft elastomer materials utilized in this study can serve as packaging materials for various sensor and actuator structures. When integrated with hydrogel electronic skin, this technology holds promise for enabling a customizable pixelated interface for emotional management in wearable devices. Interdisciplinary collaboration with experts in neuroscience and psychology could further expand the technology's potential applications in domains such as depression treatment, personalized rehabilitation, educational tools, and virtual reality—thereby contributing to the advancement of a more inclusive and barrier-free society.
Sources: https://spj.science.org/doi/10.34133/research.0714
