Prevention is better than cure

Hi-tech systems will assist drivers and help to improve road safety into the future.

Visualisation of sensor data and detected and classified objects (vehicles and pedestrians) and their trajectories in a virtual three-dimensional world.Visualisation of sensor data and detected and classified objects (vehicles and pedestrians) and their trajectories in a virtual three-dimensional world. Visualisation of sensor data and detected and classified objects (vehicles and pedestrians) and their trajectories in a virtual three-dimensional world. The perception system is used to detect possible collisions and to mitigate their effects with focus on the protection of vulnerable road users by (semi-)autonomous braking which is the objective of SP COMPOSE within IP PReVENT.

With over 42 000 road fatalities across the EU in 2004, improving safety is a top priority for European authorities. One way of working to achieve this is through the development of ‘safe’ technologies, which can help drivers to avoid – or at least mitigate – accidents. Bringing together car manufacturers, suppliers, research institutions and associations, PReVENT is a project that researches preventative safety applications. Dr Reiner Wertheimer, Manager and Project Coordinator of BMW Group Research and Technology, reveals a little more about the programme and explains his company’s involvement in the scheme.

How has the PReVENT project evolved since its launch?

PReVENT started as an initiative to promote technological progress in technical perception and actuation. However, during the project’s preparation – which lasted about one and a half years – the research focus shifted to driver assistance and active safety functions, in order to accommodate the European Commission’s 2001 initiative to halve the number of EU road accident victims by 2010.

Do you think this deadline is realistic in terms of PReVENT’s goals?

The timeframe is quite short for us to have a widespread impact on traffic: research projects typically last three years, while predevelopment and development cycles add another five. PReVENT started in 2004, meaning product launch will take place in 2012 at the earliest.

Since only around 8% of the car fleet is replaced every year, full market penetration will occur in 2024 – assuming a 100% take-up rate. Having said that, the research was not started from scratch. PReVENT benefits from components and technologies developed in earlier R&D projects, and in turn, will feed into many future research projects.

What about Intellectual Property?

The guidelines are clear. The project generally is ‘open source’, though not all of its components are. Past knowledge is provided if essential for the project, but this cannot necessarily be used by partners outside the project. New knowledge is available to those participants who contributed to developing it.

What will happen when the research ends in 2008?

Many useful applications will have been demonstrated by this deadline. But not all will translate directly into products, as industrial R&D processes are quite complex. A great deal of effort is required for predevelopment, focusing on production reliability, adequate packaging, small weight and power consumption and low production cost.

For reasons of product liability, the legal implications need to be considered, along with a detailed investigation as to how the new functions will actually work in traffic. Field-testing has been proposed in order to facilitate and accelerate the introduction of products onto the market. However, partly since accident scenarios are so diverse, large fleets of standard vehicles would be required to ensure the statistical significance of such investigations.

How is the project structured?

PReVENT is organised into groups which carry out functional and cross-functional subprojects. Four of the groups focus respectively on: safe speed and safe following distances; lateral support; inter - section safety; and collision mitigation. A fifth group looks at combined activities for example, integrating different safety functions at the level of the human-machine interface. Most subprojects are application-oriented, while a few focus on technology. None of the projects have a budget exceeding €4 million, in order to keep individual activities manageable.

Which of the subprojects has BMW been involved in?

BMW Group Research and Technology has mostly been involved in cutting-edge projects such as developing wireless local danger warning; integrating digital maps and autonomous cruise control; intersection safety; and collision mitigation through autonomous braking. Our interest in collision mitigation and pedestrian protection has also motivated us to promote the development of a distance-sensitive range camera. This allows detailed object distances to be captured in image sequences.

How does the PReVENT consortium work together?

DaimlerChrysler is PReVENT’s coordinator, while each partner contributes to a different degree – BMW’s contribution has been one of the largest. A core group of about a dozen key partners acts as the decision-making body for the project.

Collaboration has generally been good, although a few partners have struggled to deliver. At the subproject level the organisation has been quite efficient, but on the macro level, this has been more difficult. The prestige surrounding the project, as well as the large budget of around €55 million – including €30 million in EU funding – has led to more political (and, in turn, management) oversight than usual, adding to project overheads.