Question

Describe a relational database that would be useful in storing the beginning, ending and all intermediate stages for blockworld robot movements.

Answer 1

Relational Databases

Databases can be organized in many different ways, and thus take many forms. The most popular form of database today is the relational database. Popular examples of relational databases are Microsoft Access, MySQL, and Oracle. A relational database is one in which data is organized into one or more tables. Each table has a set of fields, which define the nature of the data stored in the table. A record is one instance of a set of fields in a table. To visualize this, think of the records as the rows of the table and the fields as the columns of the table.

We have been developing a paradigm that we call learning-from-observation for a robot to automatically acquire a robot program to conduct a series of operations, or for a robot to understand what to do, through observing humans performing the same operations. Since a simple mimicking method to repeat exact joint angles or exact end-effector trajectories does not work well because of the kinematic and dynamic differences between a human and a robot, the proposed method employs intermediate symbolic representations, tasks, for conceptually representing what-to-do through observation. These tasks are subsequently mapped to appropriate robot operations depending on the robot hardware. In the present work, task models for upper-body operations of humanoid robots are presented, which are designed on the basis of Labanotation. Given a series of human operations, we first analyze the upper-body motions and extract certain fixed poses from key frames. These key poses are translated into tasks represented by Labanotation symbols. Then, a robot performs the operations corresponding to those task models. Because tasks based on Labanotation are independent of robot hardware, different robots can share the same observation module, and only different task-mapping modules specific to robot hardware are required. The system was implemented and demonstrated that three different robots can automatically mimic human upper-body operations with a satisfactory level of resemblance.

Fig 1

From: Describing Upper-Body Motions for Learning-from-Observation Robots

Task recognition and state transitions. a Task recognition. Abstract task model associates one state transition with a necessary robot operation to create the transition and b state transitions and associated robot actions in the two-block world.

We define our task recognition scheme as an extension of object recognition. In the offline mode of object recognition, abstract object models are prepared and stored in a computer’s database. In the online mode, the computer associates model features with real features, identifies the corresponding abstract objects, and creates a world representation with instantiated object models. Similarly in task recognition, in the offline mode we prepare abstract task models on a computer that associate state transitions with the operations necessary to create such transitions. In the online mode, the system detects state transitions from the object recognition result and identifies an abstract task model to associate the detected state transition with an operation to achieve the transition.

Fig 2

From: Describing Upper-Body Motions for Learning-from-Observation Robots

Exploration of task domains

we demonstrated the application of this task-skill paradigm to a humanoid robot for performing a Japanese folk dance called Aizu-bandai-san (Nakaoka et al. 2007). We defined tasks for the lower body as contact states between the feet and the floor. The robot had three states: left-foot-contact, right-foot-contact, and both-feet-contact. For these states, we defined three task models: right step task, left step task, and standing task. For each task model, skill parameters were defined such as step width, step height, and waist height. See Fig. 3.

Fig. 3

Although the robot could successfully perform the Aizu-bandai-san dance and attracted considerable attention from the media and academia, defining upper-body tasks (that is, describing human poses for this purpose) has been an open issue since then.

we designed tasks for upper-body operations (that is, motions of upper-body parts) based on Labanotation which is used by the dance community to describe dances. In addition, we propose a method to extract Labanotation by means of observation. The contributions in this paper are as follows:

1. 1.Obtain Labanotation by means of observation,
2. 2.Establishment of a method to describe upper-body operations as state transitions based on Labanotation, and
3. 3.Generate upper-body operations of three actual robots based on observed Labanotation.