A tailored, electronic textile conformable suit for large-scale spatiotemporal physiological sensing *in vivo*

The study addresses the need for scalable, comfortable, multimodal, large-area physiological sensing systems beyond single-parameter, single-location wearables. Existing on-skin and textile devices are often limited by small sensing areas, single modalities, the need for nearby NFC readers or multiple batteries, fragility, lack of stretchability for skin-contact suits, and high fabrication costs. The research question is whether a tailored, stretchable, and reconfigurable electronic textile platform can provide conformal, spatiotemporal, multimodal sensing (temperature, heart rate, respiration) across large body areas with reliability, washability, and comfort suitable for daily activity. The purpose is to integrate flexible-stretchable electronics with digitally knitted, body-fitted garments to enable distributed sensor networks with minimal wiring using I2C bus architecture, ensuring intimate skin contact via controlled compression. The importance lies in enabling continuous, mobile health monitoring, sports analytics, and personalized telemedicine through a scalable platform compatible with standard textile manufacturing and garment patterning techniques.

Prior work includes soft, skin-like patches for electrophysiology, temperature, pulse oximetry, hydration, and other biomarkers, employing deformable substrates or intrinsically stretchable materials. Textile-based approaches have integrated electronics via metal-coated yarns, conductive inks, sewn functional threads, woven polyimide-based strips for humidity, temperature, pulse oximetry, and gas sensing, as well as electronic fibers. However, many are not scalable for large-area sensing or lack stretchability for skin-contact garments. Distributed sensor networks have been explored, e.g., battery-free NFC epidermal sensors for pressure and temperature mapping and studies of thermoregulation and dermatome abnormalities, but these often require proximity NFC readers, are fragile, and are less suitable for dynamic activities. Battery-powered on-skin devices can be cumbersome in multi-node deployments. Recent customizable and modular soft electronics have seldom been applied to textile systems. Variability in human form complicates smart clothing design, indicating a need for a universal, modular sensor-network platform and standardized hardware-software integration on textiles to enable industry-scale personalization.

System design and fabrication: Sensor islands and interconnects were implemented on two-layer flexible PCBs (FPCB, KingCredie) with 18 µm Cu traces and 28 µm PI substrate and outer shell. Sensor ICs (MAX30205 temperature; MPU6050 IMU) and passives were assembled and encapsulated with washable encapsulant (PE773, DuPont) and a TPU shell (TE-11C, DuPont, ~100 µm per side). Modules expose four pads (VDD, SCL, SDA, GND) and connect to reconfigurable serpentine interconnect strips with 1 mm x 4 mm pads enabling cut-to-length wiring. Temperature sensor addressing (A0–A2) was hard-wired in variants (M, M1, M2, M3) to cover up to 32 unique I2C addresses (0x40–0x5F); IMUs support two addresses (0x68, 0x69). A single I2C bus supports up to 128 addresses. Digital knitting and garment construction: Customized double-layer knit fabrics with integrated hollow channels were produced on a flat two-bed digital knitting machine (Super-J 212, Matsuya). Channel width was ~1 cm to accommodate 0.6 cm-wide modules. Interlocked patterns connected fabric layers where needed. Large fabric panels (55 x 120 cm) for front, back, and sleeves were knitted, then laser-cut and tailored with ~10% horizontal reduction to ensure compression fit. Pieces were sewn (zig-zag stitch) into a bodysuit. Modules were woven through channels, exposed via solder-tip melting for access, adhered with washable fabric glue, and interconnected to a main hub via four vertical thin copper wires routed through seams. Electronics integration and wireless: The main hub (MetaWearR, Mbientlab) includes a microprocessor, BLE, and a 3.7 V, 100 mAh Li-polymer battery (401622, HYP). Conductive snaps formed the textile-hardware interface for I2C. Total current consumption with all nodes active was ~18.6 mA, yielding ~5 h 20 min operation per 100 mAh charge. Data were streamed over BLE to a computer for logging and real-time visualization (Python Matplotlib/pygame). Mechanical and electrical testing: Interconnects (two serpentine lines, 10 x 20 mm samples) underwent uniaxial tensile tests to rupture (Instron 5943, 0.5 kN load cell, 1 mm/s) and fatigue tests (1000 cycles at 30% strain). Resistance was measured synchronously (E4980A LCR meter). FEM simulations (COMSOL Multiphysics 5.0) modeled stress distribution of serpentine interconnects (TPU hyperelastic/viscoelastic), with 30% deformation and symmetry boundary conditions. Compression pressure modeling and measurement: Fabric rigidity was derived from tensile tests on knit samples (5 x 10 cm) at 200 mm/min. Using Laplace’s law and a compression factor (CF) of 0.965 for upper limbs, garment pressures were modeled across arm circumferences and validated with a sub-bandage pressure monitor (Kikuhime, TT Meditrade). Sensor characterization: Temperature sensors (encapsulated, with and without fabric embedding) were validated on a hot plate ramping 25–50 °C (300 °C/h) against an IR camera (Optris PI 400i, 25 Hz, 40 mK sensitivity, ±2% accuracy). Calibration via linear fit (offset, multiplier) aligned sensor to IR readings; 2-D FEM heat-transfer simulations matched experimental results. IMU-based seismocardiography (z-axis, 100 Hz, ±2 g, 0.0012 m/s² precision) was acquired below the sternum and compared with commercial ECG and respiration (Zephyr BioPatch) in a resting supine subject. FIR filters were used to isolate cardiac and respiratory components. Washability and breathability: Washability was evaluated by integrating LED strips and sensor modules into textile patches, running 10 wash cycles and a continuous real-time industrial washer test (Maytag MHN33PDCWW0, delicate/knit, cold water), streaming temperature and acceleration during warm wash, rinse, drain, and spin. Breathability (WVTR) followed ASTM E96 using sealed Petri dishes with water and different fabrics (cotton, polyester/spandex, custom double-layer polyester), weighing daily at 21 °C and 50% RH. Activity study: A healthy male subject performed a treadmill protocol (2 min rest; 9 min run at 6 mph graded load; 1 min walk at 3 mph; 3 min rest). The E-TeCS streamed 30-node temperature (1 Hz) and IMU (100 Hz) data. A separate session without E-TeCS used FLIR Duo R thermography for comparison. FIR filters were designed for ECG/respiration signal processing as specified.

- Tailored E-TeCS achieved intimate skin contact with comfortable compression (~2–20 mmHg across sleeve, engineered; ~25 mmHg stated) enabling precise sensing. - Temperature sensing accuracy 0.1 °C and precision 0.01 °C using MAX30205; calibration aligned sensor to IR thermography with mean tolerance 0.2308 ± 0.0488 °C over 30–40 °C in FEM and experiments. - IMU-based mechano-acoustic sensing (MPU6050, ±2 g) captured seismocardiography features aligned with ECG, enabling heart-rate and respiration monitoring; precision 0.0012 m/s². - Distributed network capability: a single I2C bus supports up to 32 temperature nodes and 2 IMUs; prototype suit integrated 31 sensor islands (30 temperature nodes over ~1500 cm², 1 IMU at sternum) and wirelessly streamed data. - Mechanical robustness: Serpentine interconnects maintained stable resistance (single-line 0.32–0.45 Ω; module between two lines ~0.6–0.8 Ω) up to ~80% strain; rupture around 79–88% strain. Withstood 1000 cycles at 30% strain without significant electrical degradation. - FEM stress simulations indicated maximum stress at serpentine arcs (±90°); simulated tensile strength (1.8335 MPa) agreed with experimental (1.8755 MPa) at 30% deformation (<10% error). - Washability: Encapsulated modules withstood multiple washing cycles; continuous industrial wash test (34 min) showed temperature tracking of wash stages and accelerometer capturing slow/medium/fast spins; encapsulation showed no flaking/discoloration after 10 cycles. - Breathability: Despite greater thickness (1.9 mm), custom double-layer knit fabric exhibited WVTR 6.22% higher than sports polyester/spandex fabric; 31.04% lower than open-air reference. - Activity study: E-TeCS mapped spatiotemporal skin temperatures (30 nodes). During graded running, initial temperature increases were followed by decreases due to perspiration; posterior arms showed minimal trend changes. IMU detected running cadence (~174 steps/min at 6 mph) and transitions; mechano-acoustic waveforms showed resting heart rate ~108 bpm with respiration ~24 breaths/min before exercise, rising to ~156 bpm and ~54 breaths/min post-exercise. - System usability: Main hub current ~18.6 mA with all nodes active; 100 mAh battery provided ~5 h 20 min operation; detachable snaps enabled charging and maintenance.

The results demonstrate that a tailored, knitted e-textile suit can serve as a scalable platform for multimodal, large-area physiological monitoring with comfort and durability suitable for daily use. The engineered compression ensures consistent sensor-skin coupling, improving temperature accuracy and mechano-acoustic fidelity. The modular I2C architecture simplifies wiring and enables reconfiguration and personalization while maintaining low system complexity. Calibration and FEM modeling validated thermal measurement accuracy and informed design choices for sensor embedding. Electromechanical testing confirmed that serpentine interconnects retain conductivity under realistic garment strains and repeated use, and FEM stress analysis aligned with experiments, supporting design robustness. Washability tests and breathability measurements indicate practical garment care and wearer comfort. Compared with camera-based thermography, E-TeCS provides higher accuracy at discrete body locations without line-of-sight constraints and supports mobile, clothed use. The activity study illustrates the platform’s ability to capture meaningful physiological and movement metrics (skin temperature dynamics, cadence, heart and breathing rates), addressing the research aim of spatiotemporal sensing across the body in motion.

This work introduces E-TeCS, a tailored, digitally knitted electronic textile suit integrating flexible-stretchable sensor islands and reconfigurable interconnects for large-area, multimodal physiological and movement sensing. Key contributions include: (1) a scalable textile-electronics integration method compatible with high-throughput FPCB manufacturing and digital knitting; (2) a modular I2C-based sensor network enabling up to 30+ distributed nodes with minimal wiring; (3) validated thermal sensing accuracy and mechano-acoustic monitoring of cardiac and respiratory activity; (4) demonstrated mechanical robustness, washability, and breathability; and (5) real-time, wireless spatiotemporal monitoring during exercise. Future work will expand sensing modalities (humidity, pressure, optical, ultrasonic, gas, magnetic), deploy and evaluate the system outside the lab in diverse activities, and further optimize electromechanical reliability and wash durability for long-term, real-world use.

- Human testing was demonstrated on a single healthy male subject for the activity study, limiting generalizability; broader cohorts and varied body types are needed. - Thermal characterization relied on hot-plate and IR camera cross-validation and simplified FEM assumptions (e.g., no airflow), which may differ from real-world convective conditions. - The prototype integrated 31 sensor islands; while the architecture supports more, practical scaling (wiring, power management) requires further validation. - Battery capacity in the demonstrated hub provided ~5 h 20 min operation; extended continuous monitoring would require higher-capacity batteries or power optimization. - Current sensing modalities are limited to temperature and inertial mechano-acoustics; additional modalities are proposed but not yet integrated or validated. - Washability was tested under delicate cycle conditions; performance under harsher laundering and long-term garment wear requires extended studies.