Community Research and Development Information Service - CORDIS

H2020

AINARA Report Summary

Project ID: 666736

Periodic Reporting for period 1 - AINARA (Automation and INtelligence solutions for Automated Road trAnsport systems)

Reporting period: 2015-05-01 to 2016-04-30

Summary of the context and overall objectives of the project

According to a report from Navigant Research, worldwide sales of vehicles with autonomous capability will grow from zero in 2014 to 94.7 million in 2035. Without waiting for the legal framework, which is currently being implemented thanks to the efforts of CityMobil2 and several European governments, the market study conducted as part of the TAXISAT project revealed that the first wave of user demand would emerge in Europe in 2015 for new, more flexible and efficient transport services in pedestrian city centres and closed areas such as industrial sites or amusement parks. These needs are be based on the use of conventional cars but rather on “pod-type vehicles” (lightweight electric vehicles, either individual, or collective). If major automotive manufacturers invest massively in the automation of their vehicles, this is not the case of these special vehicles manufacturers.

We believe that these short and medium term markets will grow so fast that current vehicle manufacturers, Transport Network Operators vehicles and Transport Organising Authorities won’t be able to adjust their services to these new challenges as quickly as required. These organisations and companies will need key skills in robotics, intelligence and fleet management.

This is the point where we come in: meeting the emerging requirements by providing off-the-shelf ARTS solutions, instead of the highly customised engineered applications which are currently in use. The latter are incompatible with a wider uptake of the technology that requires to deal with a wide variety of environments while offering a high quality of service. To take a solid position on the emerging global autonomous vehicles market and acquire an industrial dimension, we are convinced that we must combine the best tools and practices from software engineering, with deep robotics expertise in areas such as sensing, localisation, navigation.

Our ambition is to become a leader in the provision of software solutions for vehicle automation and fleet management, for passengers and goods transportation.

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

From a very high level point of view, the project is split in two phases:
- Phase one focuses on the development of the robotisation kit including all software components and algorithms together with the definition of an appropriate validation plan, and the set up of validation tools.
- Phase two focuses on the actual validation of the software kit and the demonstration of the developed solution for a large-scale deployment on various types of vehicles and various ARTS transportation configurations

Phase one is finished, and we have developed a complete ARTS solution. To reach this objective, our effort has been concentrated on the following topics:
- Architecture Definition: we have defined a software architecture for the autonomous navigation application, including on-board modules embedded on the vehicle and fleet management services. Moreover, a thorough use-case analysis of the ARTS systems has been performed allowing us to clearly identify the different user interfaces together with their ergonomic requirements.
- Software Development: this has been our core activity. Highly skilled software developers with a robotics background have been in charge of implementing the algorithms for autonomous navigation and adapting them to new requirements. We have developed a fleet management system required to operate a fleet of vehicles applicable to different types of ARTS sites. Graphical User Interfaces have been designed and developed. We are following modern coding standards and we make use of dedicated tools to ensure code quality (code coverage, static code analysis, …).
- Software Build Chain: These tools are essential to reach our “industrialisation” objective. We have setup a development and build environment fully in the cloud (source code versioning, packages generation, automated testing, …). The software delivery process is thus highly automatised allowing us to deliver continuously updated software versions.
- Benchmarks of New Sensors and Hardware Modules: once the software architecture has been designed, we have chosen a set of sensors compatible with our algorithms. Moreover the activity of sensor testing and benchmark is continuous, since the technologies of sensors is evolving rapidly. There is always a need to evaluate and test new sensors, that could increase the capability of the vehicle. Our special benchmarking and testing efforts were centered on different cameras, Lidars and GPS receivers, that may equip future versions of autonomous vehicles.
- Definition of Evaluation Criteria: we have prepared a validation plan for the software developed, providing explicit and quantitative criteria to be met in order to ensure safety and performance.
- Testing and validation: To ensure a most efficient testing environment, we have set up a simulation environment, which completes on-site testing. We have set up a physical testing facility with more than 4 kms of roads in an urban-type environment where we have full authorisation to navigate with our testing vehicles.
- Documentation: user manuals and training documents have been prepared and cover the full software and hardware features, for installation, configuration and usage.
- Communication: we developed a lot of communication initiatives (exhibitions, workshops) and prepared relevant communication materials.

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)

General progress beyond the state of the art can be summarised by the delivery of an industrial-grade software for autonomous navigation of driverless vehicles. Our system composed of software and related sensors meets constraints of quality, safety, reliability and has a cost compatible with a large deployment.

From a technical point of view, our progress deals with the following domains:
Localisation: we have developed a 2D Lidar and a 3D visual localisation technique, that can be fused together.
Obstacle detection: we have developed an obstacle detection technique based on Lidars technology. We have characterized our obstacle detection sensors and algorithms in various environmental conditions (fog, rain, etc.)
Path planning and following: we have developed a path planning and following method adapted to the kinematic of the EZ10 vehicle
Fleet Management: a centralised server is in charge of monitoring vehicle’s status and sending mission to the vehicles on a site.

These developments have a direct impact on the market of autonomous vehicles. We are currently able to provide potential customers with a system for the autonomous navigation of their platforms. With the first demonstrations of our solution integrated on EZ10 vehicles, we prove that these systems are already operational and economically profitable.
On the long term, as these autonomous transportation systems become more and more accepted and widespread, they will profoundly change the way we move, reducing congestion, gas emission, road accidents, and transportation time.

Related information

Record Number: 194935 / Last updated on: 2017-02-17
Follow us on: RSS Facebook Twitter YouTube Managed by the EU Publications Office Top