ESPLANADE - Efficient and Safe Product Lines of Architectures eNabling Autonomous DrivE

Self-driving cars have immense potential for reducing the number of serious traffic accidents. This is because human error is far and away the most frequent cause of accidents today. For this to succeed, we need to ensure that self-driving vehicles are always safe, despite the fact that we know they will encounter situations we have not anticipated. Before our road network is ready for self-driving vehicles in all contexts, we can introduce them gradually into traffic and create trust among all interested parties throughout society.

The main purpose of the ESPLANADE (Efficient and Safe Product Lines of Architectures eNabling Autonomous DrivE) project extension is to map out this pathway. The project started in January 2017. ESPLANADE will run for three years and will involve participants from a broad consortium of Swedish organisations. RISE is the project coordinator, and the other parties are Volvo Cars, Volvo AB, Autoliv, Delphi, Systemite, Qamcom, Semcon, Comentor and KTH Royal Institute of Technology in Stockholm.

 


Project Logo: Efficient and Safe Product Lines of Architectures eNabling Autonomous DrivE 
 

Tactical safety makes traffic much safer

We can view autonomous drive functions as a next step in a trend towards intervening ever earlier in the chain of events leading to an accident. We started a few decades ago with automated functions for passive safety. A typical example is an airbag, which is designed to alleviate the effects of an accident.

During the first decade of this century, we saw a rapid development in active safety functions. These functions help drivers who find themselves in a critical situation. A typical example is an emergency braking function that helps prevent an accident where the driver does not see an obstacle in time.

With self-driving cars, we arrive at the next step: tactical safety. This means that the car, with its automated control system, actually prevents critical situations from even arising.


Make vehicle tasks more ambitious on a step-by-step basis

A condition for tactical safety being able to work is that there should not be a manual driver who is in a position to sabotage the safe planning the car undertakes for its own driving.

Only one intelligence at a time can determine the car's planned driving Both the manual driver and the auto pilot may each have a good idea of how to act in a given situation. It may however be impossible to know how the other thinks, and if we allow the driver to intervene and take over whenever they choose, there is a serious risk that precisely this action will lead to critical situations, as opposed to the contrary being the case. Furthermore, it is generally unreasonable to expect a manual driver to be able (or indeed want!) to monitor a self-driving car for longer periods. This is why we in ESPLANADE are agreed that the path going forwards should not be a gradual increase in the degree of automation, but that we should proceed directly to cars that take full responsibility for their driving.

The way to build up knowledge and create trust in self-driving cars is to make the autonomous driving task increasingly complex. Initially, we will see simpler tasks such as motorway driving and parking without a driver in the car (valet parking).

Once we have sufficient proof of their safety to be able to take full responsibility for market introduction of the cars, we can create acceptance in society and build up experience for the next step.

The ESPLANADE project aims to provide a methodology that can be applied to increasingly complex tasks so as to be able to guarantee the safe introduction of more and more self-driving tasks for road vehicles.

Extension of a pioneer project that has attracted a lot of attention

ESPLANADE is the extension of the FUSE project that has attracted a lot of attention at both national and international level. You can read more about the FUSE project here. You can also download a brochure summarising all the results.

 

Photo source: Volvo Cars

FUSE - FUnctional Safety and Evolvable architectures for autonomy

Self-driving cars can reduce the number of serious traffic accidents. For this to succeed, we need to ensure that self-driving vehicles are always safe, despite the fact that they will encounter situations we have not anticipated.Read more...

Related Information

See also

The FUSE project homepageDokumentet öppnas i ett nytt fönster

Contact Persons

Fredrik Warg

Phone: +46 10 516 54 12

RISE Research Institutes of Sweden, Phone 010-516 50 00, E-mail info@ri.se

The RISE institutes SP, Innventia and Swedish ICT have merged in order to become a stronger research and innovation partner for businesses and society.
During 2017 sp.se will be one of several websites within RISE. Please visit ri.se for more information about RISE.

Dela den här sidan: