Learning to feel better

Abstract

Sensors Human-robot interfaces require sensing technologies that capture complex biomechanical motion with high fidelity. Kim et al. developed a sensing architecture that integrates optoelectronic, microfluidic, and piezoresistive sensing into a compact soft sensor to measure multiple modes of mechanical deformation. The researchers could distinguish between stretching, bending, and compressive deformation modes from the sensor signals using threshold evaluation and machine-learning algorithms. The soft sensor was integrated into a wearable controller to remotely operate various robotic devices. Sci. Robot. 5 , eabc6878 (2020).