Soft Multi-Directional Force Sensor for Underwater Robotic Application

Papers published in Sensors, Applied Mechanics, and IEEE FLEPS 23 (see publications)

About

In this research study, we present a facile and cost-effective synthesis technique of a flexible multi-directional force sensing system, which is also favorable to be utilized in underwater environments. We made use of four flex sensors within a silicone-made hemispherical shell structure. Each sensor was placed 90∘ apart and aligned with the curve of the hemispherical shape. If the force is applied on the top of the hemisphere, all the flex sensors would bend uniformly and yield nearly identical readings. When force is applied from a different direction, a set of flex sensors would characterize distinctive output patterns to localize the point of contact as well as the direction and magnitude of the force. The deformation of the fabricated soft sensor due to applied force was simulated numerically and compared with the experimental results. The fabricated sensor was experimentally calibrated and tested for characterization including an underwater demonstration. Later, a machine learning algorithm was developed to classify the region of the applied force on the sensor. This study would widen the scope of identification of multi-directional force sensing, especially for underwater soft robotic applications. (see Publications)

Fig: Working principle and simulation of the sensor

Fig: Machine learning-based force region classification

Video: Sensor output when applied a normal force (operating underwater)