Model-Driven Variability Management in Service Robotics Based on Composition and Separation of Roles скачать в хорошем качестве

Model-Driven Variability Management in Service Robotics Based on Composition and Separation of Roles

This video illustrates the application of a model-driven method for variability management in service robotics based on a set of concrete examples in the domain of intralogistics. The general method is designed in the context of a robotics software ecosystem which is based on composition of building blocks and separation of roles. It can be used to model, compose and resolve variabilities of building blocks. The method can be applied to different levels of abstraction of building blocks: Components Skills (system of components) Tasks (system of skills) Multi-robot tasks In this video we apply the general method to multi-robot tasks. This means that either multiple robot resources and/or multiple tasks are available as variabilities that need to be matched against each other. In the examples, we always have a task problem building block (requirement) and a set of robot resources (each represented by building block models) that are potential solutions for the task problem. An important part of the general method is then to make sure that valid combinations of robot resources based on their set of provided capabilities are determined in order to fulfill the functional part of variability management at the level of multi-robot tasks. Next to capabilities, (non-functional) properties of building blocks are further decisive variabilities with respect to the matching of requirements. Dependency Variability Graph (DVG) models can be used to describe the causal relationship between such properties and other existing variabilities. Based on such models, a DVG solver is then able to determine customized bindings of active variabilities that fulfill the specified requirements to such properties.