User-Friendly Robot Environment for Creation of Social Scenarios - Foundations on Natural and Artificial Computation - page 212

The robot architecture is developed as a part of TiViPE visual programming
environment, which is a powerful integration tool.
As a particular application we feature training social skills to autistic children
[1,2,3]. In a broader sense, educators, medical therapists, psychologists, scenario
and storyboard designers, and software developers are the potential users of the
architecture. Hence, we need to take a end-user approach towards robot scenario
construction.
Several theories of controlling robots like the subsumption theory of
Brooks [4], or a three layered approach of reaction-control, sequencing, and plan-
ning, emerged [5,6]. A recent development is the ROS open source operating sys-
tem for robotics [7]. ROS runtime graph is a peer-to-peer network of processes
that are loosely coupled using the ROS communication infrastructure. ROS is
a is a distributed framework of processes that enables executables to be indi-
vidually designed and loosely coupled at runtime, and aims to support multiple
textual programming languages. Yet Another Robot Platform (YARP) also sup-
ports building a robot control system as a collection of programs communicating
in a peer-to-peer way, its main focus is on long-term software development [8,9].
Another platform is the Open RObot COntrol Software (OROCOS) with focus
on thread-safe and real time aspects between tasks [10]. Orca, initially part of
the Orocos project, is an open-source framework for developing component-based
robotic systems. It provides the means for defining and developing the building-
blocks which can be connected to form arbitrarily complex robotic systems, from
single vehicles to distributed sensor networks [11]. The authors of Orca make a
comparison between seven different robotics software systems [12], but also ad-
dress a key issue in robotics that the sheer amount of software necessary for event
the most basic competency is quite large . These factors make software reuse at-
tractive, i.e., one hopes to integrate existing software modules within a software
framework . TiViPE which is a graphical programming environment [13], em-
phasizes on the integration of existing software routines from (existing) libraries
without additional programming. TiViPE also covers the aspects of multiple pro-
cesses on multiple (embedded) computers, peer-to-peer communication, graphi-
cal programming, massively parallel (GPU) processing, and multiple operating
system support.
Because of the mentioned reasons, we build further on TiViPE visual pro-
graming environment.From the perspective of the user of such an environment,
our focus will be to create a modular structure of command blocks, as already
present in some simple solutions ( Choreagraph) with the possibility to design
new or redesign existing behaviors by a simple textual language used within the
command blocks.
To satisfy this purpose, an incremental approach, that is similar to sticking
Lego R

bricks together, is used. It implies that the modular components serve
as partial scenarios that can be modified by changing the order of usage, by
replacing, or adding of such modules. Using such an approach will enable scenario
designers, educators, and therapists to get full control over the robot.