Cyber-Physical Systems Engineering Labs 3rd Open Call for innovation projects : Safe autonomous vehicles

Date de clôture : 22 juin 2016  
APPEL À PROJET CLÔTURÉ

Overview

This call targets experiments to transfer or extend innovative approaches of realizing safe autonomous vehicles. Such systems have the potential to revolutionize areas like public transportation, manufacturing, domestic services, rescue operation, agriculture and mining. However, a pre-requisite to their widespread deployment is their compliance with high dependability requirements, including safety-critical ones.

Specific Challenge

CPSE Labs France aims to make dependability technologies affordable for companies that are developing autonomous vehicles, e.g., mobile ground robots, UAVs or intelligent cars. To serve this aim, the Design Centre provides several methods on tools as described in the following section.

The call is in direction to various categories of partners: (i) technology users, for experiments aiming to transfer and assess the above technologies by applying them to critical use cases; (ii) technology suppliers, for experiments aiming to consolidate and extend the set of methods and tools available for building safe autonomous vehicles. Experiments involving both technology users and technology suppliers are also welcome, in order to establish new partnerships for the integration of products or services for autonomous vehicles. Experiments enabling an increase of technology readiness level of results achieved by former CPSE-Labs experiments are also encouraged.

Design Centre Support

The experiment is jointly supported by the French and the Swedish Centres (with no incurred cost to the experiment). The French Design Centre provides the experiment with expertise and services and also with access to the following technology platforms

• For system dependability analysis – HAZOP-UML is a model-based safety analysis method to identify operational risks due to human-robot or robot-robot interactions. AltaRica-based analysis allows the assessment of a candidate architectural solution, with an allocation of safety levels and budgets.

• For safety enforcement – SMOF is a Safety Monitoring Framework to derive the specification of a set of safety monitors that launch safety interventions when dangerous states are detected. The synthesized strategies ensure safety while minimizing impact on the functional activity of the system.

• For rigorous software engineering – GenoM and MAUVE are two model-driven, component-based frameworks for developing robotic software functions. GenoM is compatible with mainstream robotic middleware (PocoLibs, ROS, Orocos). It has gateways to tools for formal verification at the model level (see in particular the on-going experiment to integrate GenoM and the BIP formal framework, http://www.cpse-labs.eu/ex_c1_fr_v&v.php). MAUVE is compatible with Orocos/ROS. It comes with formal verification at the code level, using code instrumentation and verification of timed execution traces.

• For Software-In-the-Loop testing – MORSE is a generic simulator for robotics based on the Blender Game Engine. The MORSE-based test method provides a framework for the generation of virtual worlds and missions, allowing robots to be tested in a variety of situations. This framework may be adapted and extended to other robotic simulators allowing software in the loop testing.

The Swedish Centre provides complementary European expertise about the technologies and markets for autonomous vehicles.

Proposals should indicate the desired types and level of support required:

• Hosting of experiment meetings.

• Training on the technology platforms.

• Collaborative Project Work to assist with focused activities that may include:

  • Scoping of use cases and demonstration artifacts o Surveys and background research
  • Technical work on the methods and tools
  • Communications support

Expected Results

The expected results depend on the category of experiments.
For experiments applying existing technologies to a new use case:

• Each transferred technology shall be illustrated on artifacts derived from the use case that are representative enough but do not raise IP or confidentiality issues. These outcomes shall be in an easily understood way and available for later demonstrations at the Design Centre.

• The experiment results shall describe what was valuable for the use case, what were the difficulties encountered, and possibly suggest future direction to resolve these difficulties.

• The results must be summarized in a publishable experiment description document.

For experiments aiming to develop new technologies or services, or to work on their integration:

• Each experiment shall provide a suitable demonstrator, which illustrates the created benefit in an easily understandable way and is available for later demonstrations at the Design Centre.

• Achievements must be compared to the situation preceding the experiment, where possible on the basis of measured values from industrial applications.

• The results must be summarized in a publishable experiment description document.

Expected Impact

• The results gained from the experiments shall be taken as starting point for the development of new products or services supplied to customers or directly transferred and applied for industrial use.