Eleven Integrated Technical Projects (ITPs) have been selected after the first cut-off date of the second RobMoSys Open Call. The following projects will receive funding from the European Union’s Horizon 2020 Research and Innovation Programme under the RobMoSys project.

Overview of the selected ITPs of the second Open Call:

Advanced Robot Simulations for RobMoSys (AROSYS)

Nowadays, simulations have become a fundamental tool used in most robotics development workflows, to design, test and validate robotics systems. Simulations already play a central role in the RobMoSys approach, however, the current simulation software used in RobMoSys (Gazebo) is lacking some fundamental features for an efficient use in the RobMoSys framework. AROSYS aims at integrating the Webots robot simulator into the RobMoSys framework to provide additional capabilities to users such as reproducibility, predictability, support of Windows, Linux and macOS, high-fidelity rendering, sensor accuracy, and stable physics engine.

 “With AROSYS we will bring cutting-edge simulation abilities to the RobMoSys framework!”

Consortium:

  • Cyberbotics: David Mansolio, Fabien Rohrer
  • EPFL: Michael Perret

Composable Models for Compliant Interaction Control (CMCI)

With the advent of highly redundant robots with model-based computed torque control, the specification of compliant behavior is increasingly important to create interactions with humans like cooperate pushing, carrying, walking, and with objects e.g. in grasping, polishing, or deburring. The idea of the CMCI project is to tackle the composition challenges that occur through the inevitable interplay of control, sensing, and the mechanism that creates the targeted compliant behavior, which is necessary to realize advanced robotics systems. To this end, the project aims at providing composable domain-specific (meta-)models to enable the specification of compliant behavior and model-transformations to synthesize the required motion control components, realizing the desired compliant interaction tasks. Both, the modeling and synthesis aspects will be realized in our workbench for modeling robot control architectures “CoSiMA”.

 

“We see the benefits of applying model-driven software engineering in the context of RobMoSys in the ability to describe advanced functional aspects such as the robot behavior for compliant interaction tasks, which heavily depend on the interplay of different (sub-)domains, and to empower non-control experts to easily make use of compliant interaction tasks in their own application use cases.”

 

Consortium:

  • Bielefeld University (UniBi): Dennis Wigand, Dr.-Ing. Sebastian Wrede
  • Technische Universität Braunschweig (TUBS): Pouya Mohammadi, Prof. Jochen Steil

Component Composition from Real-time Function Blocks

Today’s Roboticists can choose from a number of different frameworks to build modular component based applications. Yet, the software responsible for hard real-time motion control is still frequently developed as monolithic components.

The idea of COCORF is to overcome this by combining the lightweight microblx function block framework with the RobMoSys meta-modeling approach. In a nutshell, this will allow to replace monolithic components by compositions of and self-describing function blocks, thereby fostering reuse while simplifying testing and maintenance.

Additionally, connector blocks to general purpose robotics frameworks like ROS will be developed to allow straightforward integration with existing non-real-time components. To illustrate approach and technology, a reference architecture for the example of a mobile manipulation system will be developed and made available to the community.

 

“I am excited to contribute to RobMoSys because I am convinced that the RobMoSys approach of driving development by means of composable, domain specific models has the potential to overcome many of the problems faced in the development of complex systems today. In particular I believe this approach will benefit the often overlooked field of software engineering for real-time motion control systems in terms of reducing development time while improving overall quality!”


 

Consortium:

MKIO: Markus Klotzbücher

Energy-Guided Control Stacks and Robot-Software Architectures using Model-Driven Design

In EGCS, we extend the concept of energy-guarded loop controller components to the whole control stack, so, including sequence controllers and supervisory controllers.In practice this means using the energetic information – which is available through the energy guards that are developed in the “EG-IPC” ITP – of the components as status information on those components, such that higher-level control layers can infer whether the system operates safe and as expected.This contributes to the autonomy and reliability of the system, as it becomes fully aware of the physical interactions with the environment​.

With this project, we are addressing the challenge of “predictability and management of system-level properties of manipulation applications”. By following the RobMoSys model-driven composition approach, we aim to create models and meta models of a component-based energy-aware motion stack extension that will enable:

  1. Predictable system-level energetic behaviour providing passivity and thus stability and safety properties of component-based, networked systems.
  2. Responsive task execution by using information on the energetic state of the system at the skill and service levels, resulting in higher predictability of system performance.
  3. Information on the energetic state and energy usage of the system can be used to provide the human operator with valuable information on the energy that is consumed in during task execution and how that compares to expected values.

Consortium:

  • TNO: Bart Driessen
  • UTwente: Dr. Jan Broenink
  • VIRO

Formal Safety Analysis in Modular Robotic Applications

The ForSAMARA project focuses on formal verification of collaborative robotic applications against safety properties. In particular, we apply model checking methods where safety-critical situations can be identified by the result of a mathematical verification proof during the design time.

On the first side, this method is very powerful in its verification significance, but on the other side, model checking still suffers a complexity problem. Thus, behavioral and architectural models of robotic components have to be abstracted to reach a clear balance of verifiability and expressiveness in order to get valuable verification results.

Our motivation and expected industrial impact is a further step towards pushing formal verification methodologies into an industrial model-driven software and system design flow for robotic systems. Especially the feature of reflecting system safety requirements as properties and showing that they are unambiguously fulfilled in the designed system model can be very beneficial in modern I4.0 applications where complexity and application agility is steadily increasing.

 

Consortium:

  • Joanneum Research: Dipl.-Ing. Dr. techn. Michael Rathmair
  • Technische Universität Wien: Dipl.-Ing. Dr.techn. Ralph Hoch
  • PILZ: Dipl.-Ing. Christoph Luckeneder

QoS Metrics-In-the-loop for better Robot Navigation

Navigation is an essential capability in most robotic solutions, being basic in a wide range of scenarios, such as maintenance, inspection or factory intralogistics.  However, traditional navigation approaches are no longer adequate due to their little flexibility. Dealing with variability in open-ended environments requires robots to adapt themselves according to the current situation in order to achieve the required quality of service. In this sense, runtime adaptation allows moving autonomous navigation one step forward. The ambition of ​MIRoN ​is to provide a complete framework able to endow robots with the ability of self-adapting its course of action according to the external and internal context at runtime.

The ​MIRoN framework, delivered as an ​Eclipse plug-in​, will provide both ​modelling and code generation tools enabling the creation of RobMoSys-compliant systems with adaptive navigation capabilities. At design-time, the framework is intended to support the modeling of

  1. variation points​, which determine the decision space of the adaptation process, i.e., the answer to ​what ​can be adjusted. Variation points will be linked to elements in the specification of the robotic behavior (e.g., tasks and skills);
  2. contexts​, mainly expressed in terms of RoQME[1] QoS metrics; and
  3. adaptation policies, explicating how to configure the variation points depending on the current context in order to optimize relevant non-functional properties (NFP), such as safety or performance.

Expected Impact:

  • MIRoN is contributing to RobMoSys by developing a model-based framework for dealing with adaptive robot navigation. In particular, the proposal is built on the ​Flexible Navigation Stack[1] to provide a novel approach to adaptive navigation based on the systematic use of models for dynamically reconfiguring the robot behavior, defined in terms of Behavior Trees (BT), according to the runtime prediction and estimation of QoS metrics defined on NFPs.

Consortium:

  • Blue Ocean Robotics: Davide Faconti
  • UEX: Cristina Vicente Chicote
  • UMA: Antonio Bandera

Metacontrol for ROS2 systems (MROS)

A current challenge in robot software architectures in autonomous applications is to address task, contingency and system handling while performing tasks. Current solutions in ROS and ROS2 are typically based on complex, distributed logic that interweaves the three aspects, resulting in maintainability, reusability and reliability issues. The objective of MROS is to leverage the RobMoSys model-based approach at runtime, to provide a solution for ROS2 systems, based on architectural self- adaptation driven by ontology reasoning on the architecture models. The solution will be applied to the ROS2 Navigation stack, and demonstrated in two variants of the RobMoSys Intralogistics pilot, with the Thiago platform and with a Bosch consumer- product prototype. MROS will connect the ROS2 and RobMoSys frameworks, and further enable a metamodeling-sound usage of ontologies for robot software architecting.

 

“Composeable models@runtime of ROS2 software will support more adaptable and reliable autonomous robots”

 

Consortium:

  • TUDelft
  • BOSCH
  • UPM
  • URJC
  • ITU

Safety Component Composition for Robots (SAFE4CCRobot)

The SafeCC4Robot project aims to create a methodology and tool support for integrating components for robotics ensuring safety at system level. It will enable suppliers’ robotic components to be used at different robot systems while ensuring system will remain safe after the composition.

The focus is in achieving two main goals:

  • Develop and integrate safety methodological guidance within RobMoSys tools to ensure functional safety standards compliance from early design phases of the development life cycle.
  • Develop a safe-aware robotics compositional modelling assets and software supported on model-based compositional design, to leverage the reuse opportunities of using RobMoSys tools.

Design of safety compliance robotic systems methodology

Capitalize on and adapt AMASS methodology for compliance to integrate RobMoSys tools.

Integrate methodology guidance into robotics development platform. It will gather information from subsystems to correctly describe the compositional context and improve standards compliance.

  1. Methodology for safety development.
  2. Tools to support functional safety compliance for robotics systems with OpenCert tool

 

Design of safe-aware compositional robotic systems

Extend RobMoSys’ Component Development View with information regarding functional safety compliance for composition needs.

Extending component development view with information about safety assurance and contract based approached validation features.

Integrate the AMASS contract-based approach and tools to perform formal specification, validation and refinement of assumptions and guarantees under the compositional paradigm. This will be particularly used for composition of safety properties.

  1. Extension of the RobMoSys Component specification to support contract-based design.
  2. Tools to support the RobMoSys Component specification extension.
  3. Tools to support the RobMoSys Components integration validation.

Consortium:

  • Tecnalia

Skill composition with verified system properties

The goal of SCOPE is to contribute to the RobMoSys ecosystem by proposing methods and tools to enable the assessment of system-wide safety properties at the behavioral level (the “deliberative layer”) where safe autonomy becomes the key challenge. With reference to the RobMoSys meta-model for robotic behavior, the goal of SCOPE is to provide tools that analyze and derive properties of a task by composing the properties that describe its skills and the environment, and, at runtime, ensure the correct execution of a task by monitoring it and propagating anomalies detected at the level of the skills. The novelty of SCOPE is to use quantitative modelling and specification languages, so that it is possible to reason on real-time constraints, as well as resources, i.e, pre- and post-conditions for the correct execution of a skill deriving from the interaction between skills and software- hardware components sitting below the deliberative layer, plus the external environment which is directly or indirectly affected by the robot. – MOTIVATION Model-driven software engineering is the key factor to combine effective development of software with rigorous verification  techniques in robotics

Consortium:

  • Universita di Genova
  • IIT

Guidelines for Improving SmartMDSD with DDS and QoS attributes for communications

SmartDDS aims at advising designers and developers at University of Applied Sciences/Ulm (HSU) in order to assist them for including DDS in the SmartMDSD toolchain.

The main objectives are:

  1. to discuss the guidelines for the definition of QoS attributes for communications in RobMoSys;
  2. to evaluate and discuss about the prototype code generators already existing at HSU regarding communication patterns to be able to use DDS as an alternative middleware;
  3. to explore ways to help SmartMDSD developers with the design of a zero-copy gateway approach for exchanging data with already existing third-party DDS-based systems such as those developed in ROS 2 or systems which are part of smart environments using Internet of Things (IoT) devices.

Contact:

  • Jesús Martinez Cruz, UMA

Verifiable Composition of Dynamics and Control Algorithms for Robot Motion

Generating robot motion is mandatory to accomplish a plenitude of tasks including free-space motion and contact-oriented interactions. To compute desired robot motions a composition of kinematics, dynamics and control algorithms is required. The composition for non-trivial tasks with predictable properties at the system-level remains challenging even though the individual algorithms are nowadays well understood. VERICOMP aims at enabling developers to make verifiable statements about compositions in the robot motion domain. In particular, the project will focus on the non-functional requirements (i.e. timing) in addition to the functional ones, since neglecting the timing properties and requirements of a control algorithm can lead to a highly unstable, unpredictable and dangerous behavior.

“We see the benefits of applying model-driven software engineering in the context of RobMoSys in the ability to achieve verifiable and predictable compositions of motion control architectures.”

Consortium:

  • Bielefeld University (UniBi): Jan Moringen, Dr.-Ing. Sebastian Wrede
  • Bonn-Rhein-Sieg University of Applied Sciences (BRSU): Sven Schneider, Prof. Nico Hochgeschwender
AROSYS

Advanced Robot Simulations for RobMoSys (AROSYS)

Nowadays, simulations have become a fundamental tool used in most robotics development workflows, to design, test and validate robotics systems. Simulations already play a central role in the RobMoSys approach, however, the current simulation software used in RobMoSys (Gazebo) is lacking some fundamental features for an efficient use in the RobMoSys framework. AROSYS aims at integrating the Webots robot simulator into the RobMoSys framework to provide additional capabilities to users such as reproducibility, predictability, support of Windows, Linux and macOS, high-fidelity rendering, sensor accuracy, and stable physics engine.

 “With AROSYS we will bring cutting-edge simulation abilities to the RobMoSys framework!”

Consortium:

  • Cyberbotics: David Mansolio, Fabien Rohrer
  • EPFL: Michael Perret
CMCI

Composable Models for Compliant Interaction Control (CMCI)

With the advent of highly redundant robots with model-based computed torque control, the specification of compliant behavior is increasingly important to create interactions with humans like cooperate pushing, carrying, walking, and with objects e.g. in grasping, polishing, or deburring. The idea of the CMCI project is to tackle the composition challenges that occur through the inevitable interplay of control, sensing, and the mechanism that creates the targeted compliant behavior, which is necessary to realize advanced robotics systems. To this end, the project aims at providing composable domain-specific (meta-)models to enable the specification of compliant behavior and model-transformations to synthesize the required motion control components, realizing the desired compliant interaction tasks. Both, the modeling and synthesis aspects will be realized in our workbench for modeling robot control architectures “CoSiMA”.

 

“We see the benefits of applying model-driven software engineering in the context of RobMoSys in the ability to describe advanced functional aspects such as the robot behavior for compliant interaction tasks, which heavily depend on the interplay of different (sub-)domains, and to empower non-control experts to easily make use of compliant interaction tasks in their own application use cases.”

 

Consortium:

  • Bielefeld University (UniBi): Dennis Wigand, Dr.-Ing. Sebastian Wrede
  • Technische Universität Braunschweig (TUBS): Pouya Mohammadi, Prof. Jochen Steil
COCORF

Component Composition from Real-time Function Blocks

Today’s Roboticists can choose from a number of different frameworks to build modular component based applications. Yet, the software responsible for hard real-time motion control is still frequently developed as monolithic components.

The idea of COCORF is to overcome this by combining the lightweight microblx function block framework with the RobMoSys meta-modeling approach. In a nutshell, this will allow to replace monolithic components by compositions of and self-describing function blocks, thereby fostering reuse while simplifying testing and maintenance.

Additionally, connector blocks to general purpose robotics frameworks like ROS will be developed to allow straightforward integration with existing non-real-time components. To illustrate approach and technology, a reference architecture for the example of a mobile manipulation system will be developed and made available to the community.

 

“I am excited to contribute to RobMoSys because I am convinced that the RobMoSys approach of driving development by means of composable, domain specific models has the potential to overcome many of the problems faced in the development of complex systems today. In particular I believe this approach will benefit the often overlooked field of software engineering for real-time motion control systems in terms of reducing development time while improving overall quality!”


 

Consortium:

MKIO: Markus Klotzbücher

EGCF

Energy-Guided Control Stacks and Robot-Software Architectures using Model-Driven Design

In EGCS, we extend the concept of energy-guarded loop controller components to the whole control stack, so, including sequence controllers and supervisory controllers.In practice this means using the energetic information – which is available through the energy guards that are developed in the “EG-IPC” ITP – of the components as status information on those components, such that higher-level control layers can infer whether the system operates safe and as expected.This contributes to the autonomy and reliability of the system, as it becomes fully aware of the physical interactions with the environment​.

With this project, we are addressing the challenge of “predictability and management of system-level properties of manipulation applications”. By following the RobMoSys model-driven composition approach, we aim to create models and meta models of a component-based energy-aware motion stack extension that will enable:

  1. Predictable system-level energetic behaviour providing passivity and thus stability and safety properties of component-based, networked systems.
  2. Responsive task execution by using information on the energetic state of the system at the skill and service levels, resulting in higher predictability of system performance.
  3. Information on the energetic state and energy usage of the system can be used to provide the human operator with valuable information on the energy that is consumed in during task execution and how that compares to expected values.

Consortium:

  • TNO: Bart Driessen
  • UTwente: Dr. Jan Broenink
  • VIRO
ForSAMARA

Formal Safety Analysis in Modular Robotic Applications

The ForSAMARA project focuses on formal verification of collaborative robotic applications against safety properties. In particular, we apply model checking methods where safety-critical situations can be identified by the result of a mathematical verification proof during the design time.

On the first side, this method is very powerful in its verification significance, but on the other side, model checking still suffers a complexity problem. Thus, behavioral and architectural models of robotic components have to be abstracted to reach a clear balance of verifiability and expressiveness in order to get valuable verification results.

Our motivation and expected industrial impact is a further step towards pushing formal verification methodologies into an industrial model-driven software and system design flow for robotic systems. Especially the feature of reflecting system safety requirements as properties and showing that they are unambiguously fulfilled in the designed system model can be very beneficial in modern I4.0 applications where complexity and application agility is steadily increasing.

 

Consortium:

  • Joanneum Research: Dipl.-Ing. Dr. techn. Michael Rathmair
  • Technische Universität Wien: Dipl.-Ing. Dr.techn. Ralph Hoch
  • PILZ: Dipl.-Ing. Christoph Luckeneder
MiRON

QoS Metrics-In-the-loop for better Robot Navigation

Navigation is an essential capability in most robotic solutions, being basic in a wide range of scenarios, such as maintenance, inspection or factory intralogistics.  However, traditional navigation approaches are no longer adequate due to their little flexibility. Dealing with variability in open-ended environments requires robots to adapt themselves according to the current situation in order to achieve the required quality of service. In this sense, runtime adaptation allows moving autonomous navigation one step forward. The ambition of ​MIRoN ​is to provide a complete framework able to endow robots with the ability of self-adapting its course of action according to the external and internal context at runtime.

The ​MIRoN framework, delivered as an ​Eclipse plug-in​, will provide both ​modelling and code generation tools enabling the creation of RobMoSys-compliant systems with adaptive navigation capabilities. At design-time, the framework is intended to support the modeling of

  1. variation points​, which determine the decision space of the adaptation process, i.e., the answer to ​what ​can be adjusted. Variation points will be linked to elements in the specification of the robotic behavior (e.g., tasks and skills);
  2. contexts​, mainly expressed in terms of RoQME[1] QoS metrics; and
  3. adaptation policies, explicating how to configure the variation points depending on the current context in order to optimize relevant non-functional properties (NFP), such as safety or performance.

Expected Impact:

  • MIRoN is contributing to RobMoSys by developing a model-based framework for dealing with adaptive robot navigation. In particular, the proposal is built on the ​Flexible Navigation Stack[1] to provide a novel approach to adaptive navigation based on the systematic use of models for dynamically reconfiguring the robot behavior, defined in terms of Behavior Trees (BT), according to the runtime prediction and estimation of QoS metrics defined on NFPs.

Consortium:

  • Blue Ocean Robotics: Davide Faconti
  • UEX: Cristina Vicente Chicote
  • UMA: Antonio Bandera
MROS

Metacontrol for ROS2 systems (MROS)

A current challenge in robot software architectures in autonomous applications is to address task, contingency and system handling while performing tasks. Current solutions in ROS and ROS2 are typically based on complex, distributed logic that interweaves the three aspects, resulting in maintainability, reusability and reliability issues. The objective of MROS is to leverage the RobMoSys model-based approach at runtime, to provide a solution for ROS2 systems, based on architectural self- adaptation driven by ontology reasoning on the architecture models. The solution will be applied to the ROS2 Navigation stack, and demonstrated in two variants of the RobMoSys Intralogistics pilot, with the Thiago platform and with a Bosch consumer- product prototype. MROS will connect the ROS2 and RobMoSys frameworks, and further enable a metamodeling-sound usage of ontologies for robot software architecting.

 

“Composeable models@runtime of ROS2 software will support more adaptable and reliable autonomous robots”

 

Consortium:

  • TUDelft
  • BOSCH
  • UPM
  • URJC
  • ITU
SafeCC4Robot

Safety Component Composition for Robots (SAFE4CCRobot)

The SafeCC4Robot project aims to create a methodology and tool support for integrating components for robotics ensuring safety at system level. It will enable suppliers’ robotic components to be used at different robot systems while ensuring system will remain safe after the composition.

The focus is in achieving two main goals:

  • Develop and integrate safety methodological guidance within RobMoSys tools to ensure functional safety standards compliance from early design phases of the development life cycle.
  • Develop a safe-aware robotics compositional modelling assets and software supported on model-based compositional design, to leverage the reuse opportunities of using RobMoSys tools.

Design of safety compliance robotic systems methodology

Capitalize on and adapt AMASS methodology for compliance to integrate RobMoSys tools.

Integrate methodology guidance into robotics development platform. It will gather information from subsystems to correctly describe the compositional context and improve standards compliance.

  1. Methodology for safety development.
  2. Tools to support functional safety compliance for robotics systems with OpenCert tool

 

Design of safe-aware compositional robotic systems

Extend RobMoSys’ Component Development View with information regarding functional safety compliance for composition needs.

Extending component development view with information about safety assurance and contract based approached validation features.

Integrate the AMASS contract-based approach and tools to perform formal specification, validation and refinement of assumptions and guarantees under the compositional paradigm. This will be particularly used for composition of safety properties.

  1. Extension of the RobMoSys Component specification to support contract-based design.
  2. Tools to support the RobMoSys Component specification extension.
  3. Tools to support the RobMoSys Components integration validation.

Consortium:

  • Tecnalia
SCOPE

Skill composition with verified system properties

The goal of SCOPE is to contribute to the RobMoSys ecosystem by proposing methods and tools to enable the assessment of system-wide safety properties at the behavioral level (the “deliberative layer”) where safe autonomy becomes the key challenge. With reference to the RobMoSys meta-model for robotic behavior, the goal of SCOPE is to provide tools that analyze and derive properties of a task by composing the properties that describe its skills and the environment, and, at runtime, ensure the correct execution of a task by monitoring it and propagating anomalies detected at the level of the skills. The novelty of SCOPE is to use quantitative modelling and specification languages, so that it is possible to reason on real-time constraints, as well as resources, i.e, pre- and post-conditions for the correct execution of a skill deriving from the interaction between skills and software- hardware components sitting below the deliberative layer, plus the external environment which is directly or indirectly affected by the robot. – MOTIVATION Model-driven software engineering is the key factor to combine effective development of software with rigorous verification  techniques in robotics

Consortium:

  • Universita di Genova
  • IIT
SmartDDS

Guidelines for Improving SmartMDSD with DDS and QoS attributes for communications

SmartDDS aims at advising designers and developers at University of Applied Sciences/Ulm (HSU) in order to assist them for including DDS in the SmartMDSD toolchain.

The main objectives are:

  1. to discuss the guidelines for the definition of QoS attributes for communications in RobMoSys;
  2. to evaluate and discuss about the prototype code generators already existing at HSU regarding communication patterns to be able to use DDS as an alternative middleware;
  3. to explore ways to help SmartMDSD developers with the design of a zero-copy gateway approach for exchanging data with already existing third-party DDS-based systems such as those developed in ROS 2 or systems which are part of smart environments using Internet of Things (IoT) devices.

Contact:

  • Jesús Martinez Cruz, UMA
VeriComp

Verifiable Composition of Dynamics and Control Algorithms for Robot Motion

Generating robot motion is mandatory to accomplish a plenitude of tasks including free-space motion and contact-oriented interactions. To compute desired robot motions a composition of kinematics, dynamics and control algorithms is required. The composition for non-trivial tasks with predictable properties at the system-level remains challenging even though the individual algorithms are nowadays well understood. VERICOMP aims at enabling developers to make verifiable statements about compositions in the robot motion domain. In particular, the project will focus on the non-functional requirements (i.e. timing) in addition to the functional ones, since neglecting the timing properties and requirements of a control algorithm can lead to a highly unstable, unpredictable and dangerous behavior.

“We see the benefits of applying model-driven software engineering in the context of RobMoSys in the ability to achieve verifiable and predictable compositions of motion control architectures.”

Consortium:

  • Bielefeld University (UniBi): Jan Moringen, Dr.-Ing. Sebastian Wrede
  • Bonn-Rhein-Sieg University of Applied Sciences (BRSU): Sven Schneider, Prof. Nico Hochgeschwender