Academic Journals Database
Disseminating quality controlled scientific knowledge

Robust Behavior and Perception using Hierarchical State Machines: A Pallet Manipulation Experiment

Author(s): Ramón Cintas | Luis Jesús Manso | Luis Pinero | Pilar Bachiller | Pablo Bustos García de Castro

Journal: Journal of Physical Agents
ISSN 1888-0258

Volume: 5;
Issue: 1;
Start page: 35;
Date: 2011;
Original page

Keywords: Physical Agents | robots | Computer Vision and robotics

Interacting with simple objects in semi-controlledenvironments is a rich source of challenging situations for mobilerobots, particularly when performing sequential tasks. In thispaper we present the computational architecture and resultsobtained from a pallet manipulation experiment with a real robot.To achieve a good success rate in locating and picking the pallets aset of behaviors is assembled in a hierarchical state machine. Thebehaviors are arranged in such a way that the global uncertaintyof the task is progressively reduced when approaching the goal.To do so, actions are generated in each stage that increase theconfidence of the robot of being in that particular relation tothe world. In order to set up this experiment, it is required anon-trivial set of working senso-motor behaviors. We build onthis set to design and test a pallet moving task in which therobot has to locate, approach, obtain the pose, pick up and,finally move the pallet to its target position. The only sensorysources of information available to the robot are a binocularvision system and its internal odometry. To carry out this task wehave equipped a RobEx robot with a 1 DOF forklift and a 4 DOFbinocular head. We present the conceptual and computationalmodels and the results of the experiments in a real setup.
Affiliate Program      Why do you need a reservation system?