Learning efficient haptic shape exploration with a rigid tactile sensor array
This work addresses the underdeveloped skill of haptic exploration in robotics, which is crucial for robots to interact with and recognize objects, though it is incremental by combining existing insights from human behavior and visual attention models.
The paper tackled the problem of enabling robots to perform efficient haptic shape exploration using a rigid tactile sensor array, achieving near 100% success in object contour exploration with four different objects. It developed a novel architecture that learns a generative model of haptic exploration in a simulated 3D environment by optimizing perception-action loop components through multi-module neural network training.
Haptic exploration is a key skill for both robots and humans to discriminate and handle unknown objects or to recognize familiar objects. Its active nature is evident in humans who from early on reliably acquire sophisticated sensory-motor capabilities for active exploratory touch and directed manual exploration that associates surfaces and object properties with their spatial locations. This is in stark contrast to robotics. In this field, the relative lack of good real-world interaction models - along with very restricted sensors and a scarcity of suitable training data to leverage machine learning methods - has so far rendered haptic exploration a largely underdeveloped skill. In the present work, we connect recent advances in recurrent models of visual attention with previous insights about the organisation of human haptic search behavior, exploratory procedures and haptic glances for a novel architecture that learns a generative model of haptic exploration in a simulated three-dimensional environment. The proposed algorithm simultaneously optimizes main perception-action loop components: feature extraction, integration of features over time, and the control strategy, while continuously acquiring data online. We perform a multi-module neural network training, including a feature extractor and a recurrent neural network module aiding pose control for storing and combining sequential sensory data. The resulting haptic meta-controller for the rigid $16 \times 16$ tactile sensor array moving in a physics-driven simulation environment, called the Haptic Attention Model, performs a sequence of haptic glances, and outputs corresponding force measurements. The resulting method has been successfully tested with four different objects. It achieved results close to $100 \%$ while performing object contour exploration that has been optimized for its own sensor morphology.