Engineering advanced robots with a rich set of capabilities is a complex and highly interdisciplinary task. Many developers with different backgrounds and specialization as well as domain-specific terminology and technological expertise are required to work collaboratively on robot skills and their composition to eventually build systems that perform in the real world. Hence, a coherent design, implementation, and integration of multiple domain-specific software artifacts that provide the application-specific capabilities is required.
Model-driven software development (MDSD) provides an efficient methodology that addresses many of these challenges already at design time utilizing and integrating multiple domain-specific models. While the application of MDSD for the engineering of advanced robotics systems is conceptually promising, the interoperability, composability, and reusability of developed domain-specific languages and resulting models are challenging.
In this talk, we will review the requirements and potential of language modularization and composition for advanced robotics. We will introduce a recently developed language composition approach for component-based robotics systems, which supports reuse, extensibility, and refinement of domain-specific languages as well as code generation. We present use cases for exploiting modularization and composition from different research projects, i.e. the CogIMon H2020 EU on compliant human-robot interaction. Concluding, we will discuss our experiences using the JetBrains Meta-Programming System as a state-of-the-art language workbench from a language developer’s perspective as well as from a language user’s perspective, i.e. the perspective of the roboticist supported by our set of domain-specific languages in the modeling environment.
Sebastian Wrede is head of the research group on Cognitive Systems Engineering established in 2009 by the Cluster of Excellence in Cognitive Interaction Technology CITEC (EXC 277) and Bielefeld University’s Faculty of Technology. The group aims at improved engineering methods and software architectures to support the modeling, integration, development and analysis of cognitive systems with advanced human-robot-interaction capabilities. Dr. Wrede is principle investigator in different (inter-)national projects such as the H2020 CogIMon EU project on cognitive interaction in motion and project coordinator of the BMBF project FlexiMiR in the framework of the leading-edge cluster on intelligent technical systems (it’s owl) exploring human-robot interaction and model-driven software engineering for flexible assembly systems in Industry 4.0 environments.