Rodents and Felidae whiskers exhibit remarkable sensitivity, capable of detecting both external stimuli such as airflow or physical contact and internal cues like vibrations or self-generated movement. Although engineered whiskers employ various transduction mechanisms, their functionality proves broadly applicable, encompassing tasks such as limited-range sensing of contact or non-contact interactions, surface exploration, obstacle detection, and manipulation. Despite these capabilities, utilizing whisker sensors presents identifiable challenges. One aspect we investigated is their potential for perceiving multidimensional textures. Employing a custom-designed, cost-effective whisker sensor alongside an accelerometer, we successfully categorized 18 levels of roughness, including those as low as 2.5μm, thus demonstrating the sensor's ability to provide multimodal input for textured perception. Additionally, we addressed the challenge of separating responses stemming from intrinsic and extrinsic inputs by employing a frequency domain adaptive filter (FDAF) and proposing a base-vibration response model (BVRM). The BVRM demonstrated superior performance compared to FDAF alone through simulations and experiments. Moreover, we emphasize the significance of dynamic damping in whisker sensors to enhance their effectiveness in determining contact points and mitigating vibration effects. Finally, we address the uncertainties inherent in sensor measurements, particularly those impacting motion planning and localization in robotics. Through the utilization of virtual sensor models of whisker sensors, preimages, and sensor fusion techniques, we aim to address these uncertainties, which are crucial for achieving robust sensing and tracking in robotic systems.