Intelligent driving TH« TNO I Innovation for live

Transcription

1 Intelligent driving TNO I Innovation for live TH«Intelligent Transport Systems have become an integral part of the world. In addition to the current ITS systems, intelligent vehicles can make a significant contribution to improve traffic flow, traffic safety and the environment.

2 From idea to implements of intelligent vehicles an< Recent years have witnessed considerable change in respea of vehicles, in-car systems and infrastructure. Vehicles used to have no contact with their surroundings and all the communication signals were visual and targeted at the driver. The situation is quite different now and change will continue to an even greater extent. The car vnll be connected to and communicate with its surroundings. Up till now cars have maintained maximum distance from each other. In the near fiiture the distance will be minimized to enable more efficient driving, resulting in platooning. Ultimately, road traffic will be composed of cars that automatically communicate vnth each other, their surroundings and roadside systems. This will change the role of the driver. The development of technology, acceptance and approval will be a gradual process from generation to generation. TNO has indentifiedflve generations (see illustration). To develop and implement each generation several steps have to he taken. TNO has the knowledge, models, simulators, tools and technology to take these steps in cooperation with other parties. The five generations Generation 1 Inform and influence Generation 2 Support and correct Generation 3 Single lane auto pilot Generation 4 Cooperative driving Generation 5 Fully automatic driving TNO as developer and evaluator Many players are involved in the overall performance of intelligent vehicles and cooperative systems, each with its own responsibility and objectives. For the development and implementation of products and services, considerable R&D and testing is necessary. Developing intelligent vehicle systems is also a complex, time-consuming and cost-intensive process. With the pressure of tight budgets, system developers tend to create physical prototypes early in the development process so it is important to spend time and effort on road testing. However, in many scenarios road testing is too dangerous to verify in real life. Therefore TNO has created a set of tools and facilities based on a design methodology that comprise the whole development chain and can be tailored to actual R&D needs. It is designed to support stakeholders in the development, implementation and evaluation of new systems. The advantages are: - Cost reduction Faster development Reproducible scenarios (controlled and safe circumstances) Easy use of research data User acceptance Governmental approval

3 Jtion: TNO as developer and evaluator j cooperative driving Models and simulation Models and simulation are the first step in conceptual design of new systems. TNO has several simulation tools such as PreScan, where every vehicle aspect is modelled in detail, MARS, which enables the simulation of tens of vehicles, and ITS Modeller, which can model the traffic situation of a complete city. PreScan PreScan can model and test any intelligent vehicle system in every scenario. A vehicle can actually sense its surroundings and, based on the control algorithms used, react to them. Detection technologies in PreScan are radar, laser, camera, GPS and 'car to car' and 'car to infrastucture' communication systems and the simulations are based on real physics, which guarantees reliable results. While intelligent vehicle system developers can work independently with PreScan, TNO can help customers to complete specific needs and tasks or finish entire projects as well as create tailor-made solutions using PreScan. MARS This is a multi-agent real-time high fidelity simulation tool used for the development, testing and vahdation of subsystems of cooperative transportation systems. MARS technology serves as a foundation for a new breed of tools, which are capable of very accurate, real-time simulation and evaluation. The complex interaction of objects (communication, driver models, sensor algorithms, etc.) in the world model (i.e., vehicles, sensors, actuators) can be simulated and the system or network tested in relation to robustness and dependability or in terms of different kinds of hardware and software. ITS Modeller This package fulfils the need for a modelling environment in which intelligent cooperative vehicle-infrastructure systems can be modelled, tested and evaluated for their impact on traffic flow, safety and the environment. The ITS modeller can be used during the design process of new ITS system concepts as well as in a later stage, for example, for detailed impact assessment of fully developed ITS systems. PreScan is used around the world development and evaluation of intelligent vehicle systems. One example is a car manufacturer that PreScan for the development of a lane keeping system. Our customer used PreScan to build a library of scenarios for which the lane keeping system should support the driver. A camera system was modelled and attached to the vehicle driving in each of the scenarios. Using the simulated camera images, a computer vision algorithm was developed for recognition of lane markings. Based on the output of the computer vision algorithm, the control algorithms for the lane keeping system were developed. Extensive robustness analysis was performed in PreScan, by ^m simulating different weather conditions, presence of other vehicles and sensor sor mis-alignment.

4 TNO has tested extensively a Lane Departure Warning Assistant (LDWA) in our truck-driving simulator for The I Dutch Ministry of Transport, Public Works and Water Management. The focus was to study the potential of this system for narrow lane driving. The results showed that the LDWA corrects the driving behaviour whenever the driver is distracted during the journey, improving performance, particularly in narrow lanes. But it comes at a price; using the LDWA demands greater efforl from the driver. Research results like these enable the Ministry and other stakeholders involved to develop their policy on driver support systems DESDEMONA sophisticated motion platform Simulators In the task of coupling new driver support systems and infrastructure to the end user, TNO investigates traffic behaviour using driving simulators and the most suitable methods and techniques, from mathematical models to road image observations. In this, it is vital that the right balance is achieved between reality and data interpretation. Driving simulator research offers a good relationship between these two aspects and provides many benefits. The three TNO driving simulators have their own capacities. DESDEMONA DESDEMONA is a unique simulator able to generate constant G forces. This sophisticated motion platform is used for various fields of motion research like disorientation (terrain driving and rollovers, for instance), motion cueing and motion observation. Fixed-base driving simulator The fixed-base driving simulator is suited to research questions involving responses to navigation system recommendations, for example. Moving-base driving simulator This driving simulator, positioned on a motion platform, is able to test all kinds of vehicle models as well as new concepts like steer-bywire, adaptive cruise control and lane-keeping systems.

5 Controlled environment At some point during the development process of an inteuigent vehicle system, it is necessary to test real hardware. The results must be accurate and the experiments must be performed safely. TNO has one of the best equipped test facilities in the world to perform such tests. VeHIL The Vehicle Hardware-in-the-Loop laboratory is a unique facility in which intelligent vehicles equipped with sensors and controllers can be subjected to many different traffic scenarios. The vehide-under-test is placed on a roller bench that is able to simulate road behaviour up to 250 km/hr with realistic braking and acceleration. The motions of other road users with respect to the test vehicle are represented by highly dynamic automatically guided vehicles that are designed to perform every conceivable two-dimensional course. Tests in VeHIL enable customers to benefit from a level of efficiency, accuracy, and repeatability that cannot be achieved when testing on the test track or on the open road. Accident Prevention Systems for ere is a general consensus that he; vehicle traffic is vital to the economy and prosperity of the Netherlands. The percentage of trucks on Dutch roads currently stands at 15%, a figure that is expected to continue increasing significantly. Together with the Ministry and other parties, TNO is conducting a comprehensive FOT (Field Operational Test), a large-scale pilot aimed at reducing accidents, improving safety and boosting traffic circulation. The FOT will involve approximately vehicles and will test five different systems that will not only help to prevent accidents involving trucks but also serve as a registration system records driver behaviour. Vehicle Hardware-in-the-Loop laooratory

6 Intelligent driving Test bed TNO's ambition is to develop a test bed for intelligent transport systems and cooperative systems. Such a unique R&D facility allows controlled development, testing and validation of new applications. Single applications heading towards integrated systems and simulation environments gradually moving towards controlled testing and field operational tests. The testing and validation of the integrated systems especially requires a test bed that focuses on robustness, rehability and dependability of safety related applications. This test bed is the missing link between R&D and implementation / deployment and will create a platform for open innovation and joint R&D by all relevant stakeholders. Deployment Cost-benefit analyses are central to determining policy and decisions on the introduction of intelligent vehicle systems and services. TNO develops Business Cases that establish the impact on safety, traffic flow and the environment, and thus the ultimate costs and benefits. The complexity involved in introducing cooperative systems and ITS (deployment) is increasing drastically. To be able to manage this process and enable the gradual introduction of ITS, TNO has developed tools that incorporate stakeholder analysis, risk analysis, business modelling, organisation architecture and policy as well as legal aspects. In the Safespot project a Safety Margin Assistant is being developed to warn the driver of dangers on and near the road. TNO is responsible for the non-technical aspects of introduction and making a significant contribution in the deployment scenarios, business models and cost-benefit analysis. Interactive consultation with the respective stakeholders is helping TNO work on scenarios for the gradual introduction of this intelligent vehicle system. Knowledge of testing and evaluation throughout TNO is brought together in this project. Mobility The domain of mobility is where TNO clusters its vehicle engineering expertise, broad experience of ICT applications and knowledge of both driver behaviour and traffic systems in a social context, including habitability and spatial issues. With the expertise and experience of around 400 professionals, TNO is able to offer good advice and smart products in which policy, behavioural and technological aspects are integrated. If you have any specific questions regarding mobility, just get in touch with: TNO Van Mourik Broekmanweg 6 RO. Box AA Delft The Netherlands T F mobility@tno.nl tno.nl/mobllity Contact person: RH. (Frank) Hagemeier MSc. T E frank.hagemeier@tno.nl Intelligent vehicle systems (courtmy of PReVEm-iNJERSAFEt TK#