This project aimed to create bio-realistic whisker sensors to explore how animals such as rats and harbor seals interpret tactile (touch) information from their whiskers. The project was highly interdisciplinary and its intellectual merit involved at least four disciplines.

Neuroscience: A large open question in neuroscience is how the brain combines movement and sensory information to enable animals, including humans, to perceive objects through the sense of touch. Many animals use their whiskers for touch exploration, so the whisker system is often used to investigate this important question. However, it is simply not possible to record the motion of all whiskers simultaneously during naturalistic whisking behaviors. The artificial whiskers and robotic whisker sensors developed in this project enabled us to examine mechanical signals that could not be measured in the real animal and begin to develop models of the neural processing that enables touch perception.

Material Science: We developed a novel "fiber drawing" manufacturing process to produce tapered artificial whiskers that are successful mechanical and geometric mimics of biological rat vibrissae. Tests of these whiskers showed substantial improvement over previous non-tapered filaments and the ability to predict contact point locations with a median distance error of less than half a centimeter. While developing this manufacturing process we also studied how the properties of a material are influenced by the distribution of small nano and microscale sub-phases within different zones, and the impact of the material cooling after the drawing process. Finally, we wrote a review paper that describes best practices to accurately characterize complex materials using atomic force microscopy.

Soft Robotics and MEMS sensing: We engineered three different whisker sensing systems. The first whisker sensor was designed using microfabrication resulting in a sensor the same size scale of a rat's whisker. While we were able to embed real rat whiskers in a sensor of this size, the follicle design provided limited information about the whisker deflections. The second approach took advantage of inexpensive magnetic field sensors. We attached magnet to the base of a 5x scaled-up whisker manufactured with our novel fiber drawing method, and then attached the base to a spring suspension. This allowed us to measure the moments applied to the whisker and enabled sensors that could be assembled into modular arrays. We could sense whisker deflections with minimal error. The final sensing approach used computer vision along with up to six whiskers embedded in an elastomer membrane. This approach was limited in bandwidth by the frame rate of the camera, but by using the camera to monitor the motion of both the whiskers and the membrane, we were able to distinguish between different stimuli on the whiskers including touch, airflow, and inertial motion.

Fluid dynamics: Animals use their whiskers not only for touch sensing but also to sense fluid flow. We investigated several different mechanisms for tactile and airflow sensing at small scales and will continue to work toward robot integration. In addition, we performed novel theoretical analyses that investigate how fluids interact with flexible whiskers when the airflow speed is very low.

Broader impacts of the project included outreach to K-12 populations in bio-inspired robotics. Over fifteen undergraduate researchers contributed to the success of this project. Whisker-based robots and sensing devices were presented in multiple public venues including Chicago's Museum of Science and Industry.

Last Modified: 01/08/2023
Modified by: Mitra J Hartmann