ORIGINAL ARTICLE ART 2 ool: a model-driven framework to generate target code for robot handling tasks E. Estévez 1 & Alejandro Sánchez García 1 & Javier Gámez García 1 & Juan Gómez Ortega 1 Received: 23 May 2017 /Accepted: 2 April 2018 # Springer-Verlag London Ltd., part of Springer Nature 2018 Abstract Nowadays, robotic manipulation tasks are present in modern production industries, making robotics a decisive discipline in the industrial sector. Additionally, in a short period of time, handle robots will be also become essential in daily life. There is an increase in demand for applications for handle robots with software requirements such as reusability, flexibility, and adaptability. Unfortunately, the current lack of standardization of hardware and software platforms hinders the fulfillment of these require- ments. Hence, it is necessary to define a methodology that provides guidelines to design, implement, and support at runtime of such types of applications. This work explores the advantages of Model Driven Engineering (MDE) in the design and develop- ment of tasks performed by handle robots. Concretely, the authors present the ART 2 ool (Arm based Robotic Tasks modeling Tool), a MDE framework, which is very useful for application domain experts, because it guides them along the design of the application functionality, abstracting from the emerging techniques. Besides, the proposed framework supports an automatic code generation by Mode to Text transformation techniques for component-based and ROS communication middleware, achiev- ing the requirements mentioned previously. Keywords Handle robots . Model Driven Engineering . ROS—Robotic Operating System . OROCOS 1 Introduction Nowadays, robotic manipulation is a decisive discipline in manufacturing and service industries. In fact, there is an increase in investment by public institutions in these industries in order to encourage innovation, economic growth, and job creation. Hence, initiatives such as [1–4] promote the reusability, integra- tion, flexibility, and optimization of industry processes, which are the major requirements demanded by modern production facilities due to the continuous changing market demands. Unfortunately, the fulfillment of previous commented re- quirements in handle robot-based applications is a very com- plex issue because of (1) the great variability of robots in the market for different purposes and execution platforms, (2) the proprietary solutions due to missing standards, and (3) the lack of interoperability between those tools involved in the development cycle of the tasks performed by robots in pro- duction processes. Independently of the task, every handling robot application is composed of distributed, heterogeneous software compo- nents (i.e., sensors, processing algorithms, and controllers) interacting in a highly dynamic, uncertain environment. Nevertheless, the integration and collaboration among these modules are not easy because of the lack of standards. In fact, although many elements, such as sensors, actuators, auxiliary elements, and tools, need to be added to a robot to make it more flexible and adaptable, their integration and collabora- tion are not easy because the followed software development methodology does not keep reusability in mind. In order to ensure meeting these requirements, hardware and software platforms should allow developers to cope with complexity imposed by applications themselves: hardware, software, time requirements, and distributed computing envi- ronments. In this context, this work explores the advantages delivered by the use of Model Driven Engineering (MDE) [5] to provide support to the development cycle of applications based on handling robots. In recent times, the MDE discipline is being introduced in the robotics field [6, 7]. Hence, for instance, [8] shows the * E. Estévez eestevez@ujaen.es 1 Departamento de Ingeniería de Electrónica Automática, EPS de Jaén, Jaén, Spain The International Journal of Advanced Manufacturing Technology https://doi.org/10.1007/s00170-018-1976-z