Resumen de Perception-based learning for fine motion planning in robot manipulation
- español
El problema de la incertidumbre en robotica es ineludible en el mundo real, En aplicaciones de movimiento fino, que involucran distancias cortas y contactos, las tareas son muy difíciles de modelar y el entorno no siempre es conocido previamente. Esta tesis describe una arquitectura de aprendizaje de tareas de movimiento fino para robots. El aprendizaje es autónomo, mediante repetición de experiencias, asumiendo que existe incertidumbre y que el robot se guía principalmente por sus sensores, en especial de fuerza. El método está basado en diferentes técnicas de redes neuronales y aprendizaje por refuerzo. Se incluyen simulaciones de diferentes tareas para explicar los aspectos del proceso de aprendizaje. La arquitectura se implementa en un robot real dotado de un sensor de fuerza, que aprende la tarea de inserción de piezas prismáticas con varias secciones diferentes.
- English
Robots must successfully execute tasks in the presence of uncertainty.
The main sources of uncertainty are modeling, sensing, and control. Fine motion problems involve a small-scale space and contact between objects.
Though modern manipulators are very precise and repetitive, complex tasks may be difficult --or even impossible-- to model at the desired degree of exactitude; moreover, in real-world situations, the environment is not known a-priori and visual sensing does not provide enough accuracy.
In order to develop successful strategies, it is necessary to understand what can be perceived, what action can be learnt --associated-- according to the perception, and how can the robot optimize its actions with regard to defined criteria.
The thesis describes a robot programming architecture for learning fine motion tasks.
Learning is an autonomous process of experience repetition, and the target is to achieve the goal in the minimum number of steps. Uncertainty in the location is assumed, and the robot is guided mainly by the sensory information acquired by a force sensor.
The sensor space is analyzed by an unsupervised process which extracts features related with the probability distribution of the input samples. Such features are used to build a discrete state of the task to which an optimal action is associated, according to the past experience. The thesis also includes simulations of different sensory-based tasks to illustrate some aspects of the learning processes.
The learning architecture is implemented on a real robot arm with force sensing capabilities. The task is a peg-in-hole insertion with both cylindrical and non-cylindrical workpieces.
