In a paper published in the journal Nano-Micro Letters, researchers presented a multimodal electronic skin (e-skin) integrated with artificial intelligence (AI) designed to enhance rescue robots' efficiency in post-earthquake scenarios.
Schematic illustration and photographs for the properties and applications of multimodal e-skin. a E-skin based on PVA-CNF organohydrogel can not only sense traditional temperature, humidity, and pressure information but also perceive the proximity of objects and NO2 in the environment, beyond the capacity of natural skin. b Combination of multimodal e-skin and the DNN model enables rescue robots to accurately identify buried people at earthquake disaster sites. Furthermore, the integrated NOx monitoring system, comprised of the multimodal e-skin and the gas alarm circuit, endows the intelligent robot with the function of monitoring toxic gases. c (i) Photograph of 50% tensile strain of multimodal e-skin. (ii) Photograph of the transparent multimodal e-skin attached to human skin. (iii–vi) Photographs show the e-skin being attached to a joint and following its movement without detaching. d Capability radar comparing the performance of the state-of-the-art e-skin. Image Credit: https://link.springer.com/article/10.1007/s40820-024-01466-6
The e-skin replicated natural skin's sensing abilities while adding capabilities to detect object proximity and toxic gases like NO2. Its ecoflex and organohydrogel structure provided mechanical properties like those of human skin. This innovation enabled robots to better interact with unstable environments, accurately identify objects and human limbs, and protect trapped individuals from hazardous gases.
Background
Past work on e-skin has focused on replicating natural skin's functions, but studies have yet to explore enhancing sensory capabilities beyond this. Advances in flexible electronics have led to the development of multimodal e-skin, which mimics pressure, temperature, and humidity sensing and adds functions such as object proximity and NO2 gas detection. Integrating polyvinyl alcohol-cellulose nanofiber (PVA-CNF) organohydrogel with ecoflex provides excellent mechanical properties, sensitivity, and adaptability, making it a promising solution for improving robot interaction in complex environments. This e-skin advances robot perception, surpassing traditional sensors by offering rapid response times and high sensitivity in varied conditions.
Multimodal E-Skin Fabrication
To synthesize PVA-CNF organohydrogels, deionized water, and glycerol were mixed in a 1:1 mass ratio to create an H2O-gly solution. CNFs and PVA were added in a mass ratio 1:15:100. The resulting solution was spin-coated onto an aluminum substrate at 2000 rpm for 60 seconds and left at room temperature for 12 hours, forming a PVA-CNF organohydrogel film.
The organohydrogel film was then applied to the ecoflex layer as a temperature-sensing module, followed by another ecoflex layer to encapsulate it. Additional organohydrogel films connected to silver wires were placed on the uncured ecoflex for humidity and NO2 sensing. The final structure, cured together, created a sandwich configuration for pressure and proximity sensing.
The sensing performance was characterized by placing the e-skin in environments with various relative humidity and temperatures and measuring resistance changes with an inductance, capacitance, and resistance (LCR) meter. Pressure and proximity sensing were evaluated using a motorized mobile platform and force meter, with capacitance measured through an LCR meter. Gas sensing involved applying a direct current (DC) bias to the NO2 module and recording the reaction current at various NO2 concentrations while controlling the humidity.
Advanced Multimodal E-Skin
The AI-enhanced multimodal e-skin, based on PVA-CNF organohydrogel, replicates human skin's sensing abilities while extending functionality to detect temperature, humidity, pressure, object proximity, and hazardous gases like NO2. Comprising stacked layers of conductive tapes and organohydrogel films, it ensures accurate temperature detection and can differentiate between humidity and gas presence, broadening its application scope.
The e-skin's stretchability, flexibility, and superior mechanical properties make it highly adaptable for diverse uses. Additionally, its integration with AI algorithms and hardware circuits enhances its utility in complex environments such as earthquake disaster sites, where it can assist rescue robots in identifying trapped individuals and detecting toxic gases, thereby improving rescue efficiency and safety.
The robust performance of the e-skin's humidity and temperature sensing modules is driven by the PVA-CNF organohydrogel's ability to form hydrogen bonds with water molecules. This interaction facilitates sensitive and accurate humidity detection, with the module's response being temperature-dependent. Despite its environmental exposure, the humidity module shows no cross-sensitivity to NO2, ensuring reliable measurements.
The module is protected by an ecoflex layer for temperature sensing, isolating it from humidity interference and allowing it to respond accurately to temperature changes with high sensitivity. The temperature module's design ensures prompt responsiveness, which is crucial for real-time monitoring in dynamic environments.
Moreover, the skin's proximity and pressure-sensing capabilities are finely tuned, using capacitance principles to detect changes in distance and pressure with high precision. The proximity module, for instance, leverages the redistribution of the electric field when an object approaches, leading to measurable changes in capacitance, a mechanism supported by finite element analysis simulations.
The multimodal e-skin demonstrates exceptional stability and accuracy in practical applications, even under repeated stress and in varying environmental conditions. Based on a parallel plate capacitor design, the pressure module reliably measures applied force through changes in capacitance, with sensitivity influenced by temperature.
Long-term tests affirm its durability, maintaining performance consistency over thousands of cycles. The proximity module, capable of detecting objects at varying distances, adjusts capacitance based on the proximity of conductive targets, ensuring reliable detection in real-world scenarios.
Its design accommodates environmental variables like temperature, allowing it to perform effectively across various conditions. Overall, the multimodal e-skin's combination of advanced material science and engineering design, supported by AI integration, positions it as a significant advancement in artificial skin technology. Potential applications extend beyond human sensory replication, including critical roles in hazardous environment monitoring and robotic systems.
Conclusion
To summarize, a multimodal e-skin was developed, demonstrating excellent temperature, humidity, pressure, proximity, and NOx gas sensing capabilities, surpassing natural skin. Constructed with flexible ecoflex and PVA-CNF organohydrogel layers, the e-skin exhibited high sensitivity and accuracy, particularly in detecting extremely low NO2 concentrations.
When integrated with rescue robots, these e-skin and AI algorithms significantly improved search and rescue efficiency in earthquake ruins by enabling precise environmental object classification and real-time toxic gas monitoring. This advancement enhanced robots' information processing and interaction capabilities in complex scenarios.
AI Researchers Reveal New Method for Measuring How Much is 'Too Much' in Image Generation Models