Robot control
Tactile servoing
Definition
Tactile servoing is closed-loop robot control that uses tactile measurements to reduce error relative to a desired contact state. The feedback may represent contact pose, pressure, shear, a raw tactile image, or a learned tactile feature.
Also known as: Tactile servo control
Updated
Touch closes the motion loop
A tactile servo controller repeatedly measures contact, compares it with a target, and commands motion that reduces the difference. The target could be a fingertip pose relative to an edge, a pressure pattern associated with a stable grasp, or a stored sensor observation from a demonstration.
The idea parallels visual servoing, but the feedback comes from physical contact. It can remain informative when a hand blocks an external camera or when the relevant error is how the object presses against a finger rather than where it appears in an image.
The tactile representation shapes the controller
A controller can use engineered contact features or a learned mapping from sensor data. Pose-Based Tactile Servoing estimates the pose of a soft optical tactile sensor relative to local object features, then uses that estimate in a control loop for following surfaces and edges.
Contact deformation can also contain shear caused by tangential loading. Lloyd and Lepora combine pose and shear estimates with filtering and velocity control, demonstrating surface following and pushing tasks. The paper's results concern its TacTip sensor, learned models, objects, and task setups; they do not imply that the same controller transfers unchanged to every tactile sensor.
Contact can be ambiguous and change quickly
Different object shapes or force distributions can produce similar tactile readings. Soft sensing surfaces have hysteresis, and slip or rolling contact can alter the observation between control updates. Learned estimators may fail on materials, geometries, or loading conditions outside their training data.
Tactile servoing also requires an initial contact and usually operates over a local region. A robot still needs perception and planning to approach the object, manage collisions, and decide the desired contact. Combining tactile sensing with vision, force measurement, and robot state can reduce ambiguity, but adds calibration and timing requirements.
Sources
Related terms
Tactile sensing
Tactile sensing measures information arising from physical contact, such as contact geometry, deformation, or force. Robots use it to observe interactions at their fingers, grippers, feet, or other contact surfaces.
Visual servoing
Visual servoing uses visual measurements inside a feedback loop to control robot motion. The controller updates movement to reduce an error defined from image features or visually estimated pose.
Force control
Force control regulates the force or wrench a robot applies to its environment. It may use a robot model, measured interaction forces, or both to produce joint commands that achieve a desired contact load.
Robotic manipulation
Robotic manipulation is the use of a robot to change an object's position, orientation, or state through physical interaction. It includes grasping and moving objects as well as actions such as pushing or carrying them without a grasp.