Exploiting High Quality Tactile Sensors for Simplified Grasping
Robots are expected to grasp a wide range of objects varying in shape, weight or material type. Providing robots with tactile capabilities similar to humans is thus essential for applications involving human-to-robot or robot-to-robot interactions, particularly in those situations where a robot is expected to grasp and manipulate complex objects not previously encountered. A critical aspect for successful object grasp and manipulation is the use of high-quality fingertips equipped with multiple high-performance sensors, distributed appropriately across a specific contact surface. In this paper, we present a detailed analysis of the use of two different types of commercially available robotic fingertips (BioTac and WTS-FT), each of which is equipped with multiple sensors distributed across the fingertips' contact surface. We further demonstrate that, due to the high performance of the fingertips, a complex adaptive grasping algorithm is not required for grasping of everyday objects. We conclude that a simple algorithm based on a proportional controller will suffice for many grasping applications, provided the relevant fingertips exhibit high sensitivity. In a quantified assessment, we also demonstrate that, due in part to the sensor distribution, the BioTac-based fingertip performs better than the WTS-FT device, in enabling lifting of loads up to 850g, and that the simple proportional controller can adapt the grasp even when the object is exposed to significant external vibrational challenges.