Biomimetic chameleon soft robot with artificial crypsis and disruptive coloration skin

The study addresses the challenge of creating a practical, autonomous artificial camouflage device capable of natural, high-resolution background matching and disruptive coloration. Existing pixelated approaches face complexity and scalability issues when many small, individually addressable units with sensors and control are required. The authors aim to achieve broad visible color tuning, mechanical flexibility for wearable/soft-robot applications, rapid response, and robust operation under environmental fluctuations. They propose integrating thermochromic liquid crystal (TLC) coloration with vertically stacked patterned silver nanowire (Ag NW) heaters to superpose temperature profiles at the surface, enabling both background color matching and expression of microhabitat textures without dense lateral pixelation. They also leverage the linear, non-hysteretic temperature coefficient of resistance (TCR) of Ag NW networks for closed-loop temperature control to stabilize color and speed transitions.

Artificial camouflage has military origins and has expanded into soft robotics and electronic skins. Prior work emphasizes background matching and disruptive coloration, often requiring spatial patterning. Biological systems (e.g., cephalopods, chameleons) achieve rapid active camouflage via muscular/structural mechanisms, inspiring various artificial strategies across optical regimes. Visible-range, practical RGB camouflage remains challenging, particularly with high-resolution spatial variation. Thermochromic liquid crystals offer wide tunable reflectance via narrow temperature windows and have been used mainly for thermal sensing due to high temperature sensitivity. Flexible heaters based on Ag NW networks provide electrical/mechanical stability and patternability superior to alternatives (Au NW, Cu NW, hybrids). Previous lateral pixelation increases system complexity. Biological polymorphism indicates that a limited set of pattern modes can suffice across habitats, suggesting preselected patterns rather than arbitrary images may be adequate for many applications.

Materials and device fabrication: Long Ag nanowires (≥100 µm length, ≤100 nm diameter) were synthesized via a modified polyol one-pot method using EG, PVP, AgNO3, and CuCl2·2H2O, followed by repeated ethanol washing and redispersion. Ag nanoparticles (Ag NPs) were synthesized via a modified polyol method (AgNO3 in EG with PVP), centrifuged, cleaned with ethanol, and redispersed. Multi-layered ATACS fabrication: Colorless polyimide (cPI) varnish was spin-coated on glass and annealed up to 300 °C. Ag NP ink was spin-coated and selectively laser-sintered to form base electrodes; unsintered ink was removed. Ag NW solution was spray-coated and patterned by UV laser ablation to define heater geometries. For stacking, a cPI cover layer was deposited, vias were laser-ablated, and Ag NP ink was spin-coated and laser-sintered to form interlayer electrical connections through vias. These cycles were repeated to stack up to three patterned heaters. Finally, black acrylic ink and sprayable TLC ink were sequentially coated atop the stacked heaters. The total device thickness with one heater was ~30 µm; each cPI–Ag NW layer contributed <10 µm, with the Ag NW film itself ~100 nm thick. Patterning and interconnects: Laser ablation was used to define heater patterns (uniform, stripes, wavy; also ostrich-inspired sequences; line/dot/square microhabitats). Laser sintering of Ag NPs provided through-layer electrodes and ground lines. Electronics and control: A microcontroller (Arduino Mega 2560) acquired RGB intensities from color sensors (TCS3472) via I2C at 10 Hz per sensor and determined target temperature based on background color. Heater resistance was amplified (LM324) and filtered (software low-pass 300 Hz). Temperature estimation used Ag NW TCR; a PID controller at 30 Hz drove PWM power to maintain setpoints against disturbances. Characterization: Optical imaging, IR thermography, CIE 1931 chromaticity mapping, voltage–color/temperature calibration, power measurements, cyclic on/off stability, bending tests up to curvature κ=0.70 mm−1 and 1000 bending cycles at κ=0.65 mm−1 (resistance tracking), response-time measurements under active vs passive control and at reduced ambient temperatures (15 °C, 5 °C, −5 °C). Camouflage effectiveness was quantified via RGB histograms and 2D FFT analyses comparing ATACS outputs to patterned backgrounds. A chameleon robot was outfitted with seven flexible ATACS patches; demonstrations included static backgrounds and autonomous movement across R/G/B habitats with real-time color adaptation. Data were provided via figshare.

- Vertically stacked Ag NW heaters beneath a TLC layer enable superposition of temperature profiles, transferring the selected heater's pattern to the outer surface independent of other layers, achieving both background color and microhabitat texture expression. - Tunable visible coloration: TLC operating range (example mixture) 25.5–36 °C; representative RGB at 25.5, 28, and 36 °C. One-to-one mapping between surface temperature and color with continuous transitions. - Power and efficiency: Blue coloration maintained at ~19.4 mW/cm²; a 1.2 V, 2500 mAh AA battery could sustain 10×10 cm² active camouflage for ~92 minutes (under stated conditions). - Rapid response with feedback: Active control reduced characteristic times from 3.04/3.44/3.52 s (passive) to 0.44/0.45/0.46 s (active) for R/G/B, respectively; maintained <0.6 s even at 15 °C, 5 °C, −5 °C ambient. - Robustness to disturbances: PID control using Ag NW TCR as a proxy for temperature counteracted environmental cooling (e.g., ice proximity), preserving target color; without feedback, severe color drift occurred. - Mechanical flexibility and durability: Stable RGB coloration up to curvature κ=0.70 mm−1 under constant bias; after 1000 bending cycles at κ=0.65 mm−1, normalized resistance change remained within ~0.5% with no permanent R0 shift and no color hysteresis or degradation. - Spatial resolution and patterning: Demonstrated smallest feature size of ~0.5 mm; camouflage accuracy comparable to a 64×64 pixelated device, while using only a few input channels (e.g., 3 heaters) instead of hundreds of lateral pixels. Ostrich motion patterns and microhabitat line/dot/square textures were reproduced by activating single or combined heater layers. - System integration: Seven flexible ATACS patches (21.96–67.2 cm²) conformed to a chameleon robot; autonomous sensor-integrated ATACS (S-ATACS) retrieved local RGB background and instantly matched color during locomotion across R/G/B habitats, improving concealment versus off-state. - Materials/process advantages: Ag NW networks showed linear, non-hysteretic TCR suitable for resistance-based temperature estimation; devices fabricated via ambient, largely wet processes compatible with large areas.

The findings demonstrate that a vertically stacked heater architecture paired with TLC coloration provides a practical route to high-performance artificial camouflage without the complexity of dense lateral pixelation. By leveraging the stable, linear TCR of Ag NW networks for feedback control, the system achieves sub-second color transitions and maintains target colors under environmental disturbances. The multilayer superposition approach effectively conveys fine patterns preselected for representative microhabitats, improving natural blending through both background color matching and disruptive/texture cues while requiring only a few control channels. Mechanical flexibility and durability support deployment on moving, soft or irregular bodies. The autonomous S-ATACS validates end-to-end functionality—local sensing, decision, and control—leading to instantaneous crypsis on a mobile platform. These results address the core challenge of scalable, adaptive, and natural-looking visible-range camouflage and suggest broad applicability from defense to wearable and architectural uses.

This work introduces a biomimetic artificial camouflage system combining TLC coloration with vertically stacked, laser-patterned Ag NW heaters and closed-loop control. The approach delivers rapid, low-power, and robust color tuning with the ability to superimpose predefined fine patterns, significantly reducing system complexity compared to high-resolution lateral pixel arrays. The sensor-integrated implementation achieves autonomous, instantaneous crypsis on a moving chameleon robot. Future directions include: expanding operational temperature ranges via tailored TLC compositions or hybrid heating/cooling elements; developing pattern recognition and decision algorithms to autonomously select and combine multilayer patterns; optimizing heater densities/stack counts and sensor layouts; and exploring integration onto larger, more complex moving platforms.

- TLC temperature window limits expressible colors relative to ambient: with ambient above TLC onset (~25.5 °C for the used mixture), available spectrum narrows; at very low ambient (e.g., −20 °C), steep gradients hinder uniform coloration. Mixture selection can shift operating range but requires material changes. - Current heaters provide only heating; active cooling would require thermoelectric elements, which are challenging to integrate in flexible, vertically stacked formats. - Autonomous demonstrations excluded multilayer pattern activation due to the absence of a pattern recognition and control algorithm; current polymorphism is user-selected rather than fully autonomous. - While bending robustness is shown to κ=0.70 mm−1 and 1000 cycles, long-term environmental aging and extreme mechanical strains were not exhaustively characterized. - Thermal insulation and body–skin coupling can influence performance; demonstrations used spacers/air gaps for the robot platform.