Community Research and Development Information Service - CORDIS

H2020

RobustSENSE Report Summary

Project ID: 661933
Funded under: H2020-EU.2.1.1.7.

Periodic Reporting for period 1 - RobustSENSE (Robust and Reliable Environment Sensing and Situation Prediction for Advanced Driver Assistance Systems and Automated Driving)

Reporting period: 2015-06-01 to 2016-05-31

Summary of the context and overall objectives of the project

The vision of the RobustSENSE project partners is an autonomous vehicle capable of ensuring safe and comfortable travel for its occupants and other road users under all existing driving conditions.

Perceiving and understanding the environment and ultimately predicting the movement of objects in the vehicle’s surrounding are key factors to master automated driving and ADAS. Systems on the market today are already capable of various ADAS functions under good environmental conditions. However, in adverse weather and lightning conditions the picture changes dramatically. In situations with increased risk of accident as under harsh weather or bad lighting conditions, where ADAS systems can be particularly beneficial, human control is still needed. When system support would be particularly needed, current systems fail as sensor performance decreases drastically. Consequently, the next step to be taken is to increase the robustness of safety systems in all environmental conditions.

RobustSENSE is aiming at automated and safe mobility by enabling systems to cope with real world requirements under all environmental conditions. The RobustSENSE system introduces reliable, secure and trustable sensors and software by implementing self-diagnosis, adaptation and robustness.

One goal of RobustSENSE is a robust and reliable sensor platform for automated and autonomous driving. This platform provides enhanced sensing performance, overcoming the present environmental perception systems which fail in adverse conditions. Such an improved platform is needed to achieve the necessary reliability of highly automated and autonomous driving functions for safe operation under all driving conditions. For this, the project focuses on advanced methods for improved sensor technologies and sensor signal processing as well as innovative algorithms for sensor data fusion, scene understanding, behavioural planning, and trajectory planning. Furthermore, by implementing the resulting sensor platform architecture in demonstrators, RobustSENSE will showcase the potential the developed novel technologies will have for future driver assistance functions, which are far more robust against the influences of weather and light conditions.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

Leading and coordinating the RobustSENSE project, the focus of Work Package 1: Project Management (WP1) was at the first period to start all project activities and facilitate interactions among partners. Common processes and procedures have been established and implemented (deliverable D1.1 “Project Handbook”). Project progress is tracked for each work package on a regular basis, including monitoring of milestones and deliverables as well as tracking of resources and the overall project plan. Besides the kick-off meeting, one General Assembly, a PMT physical meeting and regular phone conferences among PMT members were organised to supervise and document project progress. The progress of technical activities as well as topics regarding administration and dissemination were regularly discussed. To harmonize activities among the technical WPs, two joint WP workshops were organized. On a day-to-day basis, WP1 supported all partners in organisational and administrative tasks.

In Work Package 2: Requirements and specification (WP2), the general architecture of the RobustSENSE platform has been defined. The architecture includes all components required for online operation and for offline validation. For each of the three defined layers of the architecture – sensors, data fusion, and planning and understanding - modules and the respective interfaces between these modules have been identified and specified. Based on those, metrics and validation plans have been derived. Metrics define for each component which measurements are relevantand are used within the validation plans which describe how validation can be performed for single components or a number of components together. The validation plans also explain how the validation should be performed and which metrics are relevant for each plan.

Work Package 3: Environment perception (WP3) participants focused on deliverable D3.1 “Specification of the modular system architecture of the environment model”, with the goal to develop a system - architecture of the environmental model. This was achieved. It was aligned with the overall architecture defined in WP2 while considering the weaknesses of existing approaches. The resulting refined architecture of the level will serve as a guideline for future implementations.

In Work Package 4: Situation understanding and planning (WP4) several steps towards a coherent scene understanding, situation prediction, behavior planning and trajectory planning were undertaken, including an algorithm performance assessment. Regular telephone conferences were established and exchange between partners was increased. The conceptual purpose of the modules was defined and the first deliverable D4.1 “Specification of situation understanding and trajectory planning modules” was delivered. Every partner has already achieved some first results of his investigations and started implementing the different parts for the coherent probabilistic modules. Additionally a first paper - was published at the Intelligent Vehicles Symposium.

Work package 5: Integration and validation (WP5) aims for integration of all resulting modules and to perform prototypical vehicle testing, to validate the enhanced level of robustness with the developed validation plans. To this end there are four tasks in WP5 of which two tasks are ongoing throughout the period 2: T5.1 “System performance assessment” and T5.2 “Sensor platform prototype”. At the first period the activities focused on the definition and first development steps of the system performance assessment module. Deliverable D5.1 collects the data related to the definition of the system performance assessment module and use cases to test it. In T5.2 the assembly of sensor platforms was started, taking into account the architecture defined in WP2 as well as the interfaces needed for the purpose of the system performance assessment module.

The initial main task in WP6 was the planning and development of the overall dissemination and exploitation strategy to be followed in RobustSENSE. At the very beginning of this period a corporate identity for RobustSENSE was established and the project website created. Relevant stakeholders and channels were identified and general information material to be used by the project partners at conferences and events was developed. All work packages were supported in disseminating first project results to the public.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

Data acquisition: First concepts have been worked out for specific radar and lidar sensor concepts. For cameras, a soil detection has been implemented.

Sensor data processing: (1) Validation will be based on sensor data plus quality metrics, quality metrics have been defined in P1. (2) The software development needed for this feature started during P1.

Information fusion: (1) This is part of task 3.5 which just has been started. (2) Quality metrics for output from data fusion to environment model defined. Fusion is able to detect ambiguity by redundancy.

Scene understanding: A concept and data collection for road condition monitoring using machine learning techniques were developed.

Situation prediction: Probability metrics for other traffic participants defined.

Behavioural planning: (1) Measures and metrics defined. Work on information fusion just started. (2) Definitions of metrics for radar and lidar sensor performance assessment have been worked out.

Trajectory planning: n.a. for P1

System Performance Monitoring: Work on sensors ongoing. Definitions of metrics for radar and lidar sensor performance assessment have been worked out.
Record Number: 193027 / Last updated on: 2016-12-16
Follow us on: RSS Facebook Twitter YouTube Managed by the EU Publications Office Top