You are here: Home » Background » Proposed Concept

Proposed Concept

The goal of this activity has been clearly detailed in the PERASPERA SRC Guidelines   and Compendium  and is the development of an open-source space robot control operating system (RCOS) that can provide adequate features and performance with space-grade Reliability, Availability, Maintainability and Safety (RAMS) properties.

An operating system is much more than simply a kernel and a set of libraries: the use of an operating system implies the use of its components at design and implementation levels, and it usually defines the set of features that will be available for the development, and a set of tools usually associated to the development environment. Normally then we refer to a “framework”, i.e. a set of tools that are used in an orchestrated manner to develop a system. To give an example, the Linux OS was developed based on the (previously developed) GNU set of tools; in fact, it is more precise to talk about GNU/Linux (instead of simply “Linux”).

Therefore, it is our understanding that ESROCOS is devoted not only to the generation of an operating system for space robotics, but to the development of a robotics framework suitable for space, in which many different tools will be used. In fact, as stated in the proposal, according to PERASPERA a RCOS system integrates both the RCOS target and the RCOS development and validation (RDEV) environment. In the rest of this document, we will use “RCOS framework” as a synonym to the RCOS system

Based on this, the consortium fully answers the above-identified aspects. Instead of starting from scratch, we propose a solution based on a combination of recognized assets that have been proved previously, and that, when properly enhanced and integrated together, will fulfil such objective. The ESROCOS proposed solution is hereafter described by next schematic architectural view.


Figure 1‑2: ESROCOS proposed approach.

The innovation behind our proposed concept lies in the following main elements:

  • Use a Model-driven approach with extended, robotic-specific modelling semantics, allowing the design of a Platform Independent Model, thereby respecting the fact that space hardware varies depending on the chosen vendor;
  • Answering the needs of future robotic needs, by being able to interact across multiple components with a mixed-criticality approach, following correct-by-construction methods;
  • Developing a space-grade system, by complying with space processors and avionics drivers, command and telemetry access services as well as following SW development space-standards;
  • Ease the use of robotics developers by including basic robotics libraries, ensuring interoperability with 3D robotics viewers, simulators and third-party libraries, guarantee advanced logging capabilities as well as identifying advanced data types (closing the gap between ASN1 and robotics);
  • Exploiting end-user experience, within the consortium, in the of state-of-the art European robotics mission (ADS is the prime contractor for the ExoMars rover and DLR is the lead institute in Europe for advanced space robotics) and non-space high RAMS applications (VTT involvement in RCOS for future nuclear reactors

To summarize, the ESROCOS solution will be built around the following main elements:

  • End-user requirements and final evaluation assisted by the expertise and know-how from ADS (planetary exploration) and DLR (orbital systems). As future users of ESROCOS they will focus on the elicitation of requirements and the evaluation of the test results.
  • Led by GMV with strong assistance of DFKI as robotics advisor. Both entities are committed to convert an existing space software development framework (TASTE) into a space RCOS inspired in an existing robotics middleware (Rock where DFKI is the main developer).
  • Provision of space-grade RAMS characteristics (correct-by-construction and separation of behaviour and models) techniques with specific emphasis on the system safety properties that depend on timing and scheduling (derived from the BIP framework developed by UGA).
  • Robot meta meta-modelling capabilities based in semantics composability being developed by KUL at the edge of the latest theoretical robotics developments.
  • Based in strong formalized languages such as AADL derived from the expertise from the TASTE co-developer (ISAE)
  • Focus on new paradigms like mixed-criticality and Time and Space Partitioning (contributed by SKYas one of the Excellence Centres in IMA recognized by ESA).
  • Interoperable with ROS 2.0 middleware and its visualization and simulation tools (contributed by Intermodalics who has already done similar integration efforts for the Orocos framework).