Paris urban city centre | rFpro

Experience driving along the famous streets of Paris’ glamorous city centre. With the many iconic locations and plentiful distractions of a modern European cityscape.  

Location

As one of Europe’s most famous cities, Paris is known for its architecture, history…and challenging traffic conditions! It also hosts a Formula E race in the city centre, so rFpro has been able to model both the city streets and its racetrack at the same time. True to this location, our model has an interesting mix of surfaces including tarmac and cobbles which will challenge vehicle dynamics models.

 

Key features

The tree-lined streets and complex road network are augmented by the traditional challenges of an urban environment. Pedestrians, cyclists, trees, prams, adverts, and other objects can all prove difficult for AI to distinguish. There are also plenty of urban distractions such as prominent roadside adverts featuring objects that are distracting to human drivers but could be misidentified as their real-life counterparts by an autonomous vehicle. They are all captured within our model and appear frequently. The model also encapsulates two unusual traffic rules: a roundabout with “give way” signage, and a junction with very challenging rules!

 

Technical features

Paris is an enormously dense urban environment with a large number of modelled objects and so the rFpro team uses the Paris model for much of its internal testing. Features such as variable parked car locations and our traffic plugins were developed to allow variability in scenarios in locations like Paris. We have developed methods of moving objects within this environment to provide a variation in scenarios that is definable by the end user.
We also model our trees with finer geometry for Paris because of their prominence in the scene to ensure that during simulation, LiDAR sensors can detect them accurately.

 

Customer benefits

Even for mere mortals, Paris is one of the more difficult locations to navigate so it is an excellent model to choose if you want to challenge your autonomous vehicles. With a combination of our accurate model and fine-grained control, we can replicate reality, from an autonomous vehicle’s point of view, for training and testing purposes. This is one of the most challenging driving locations in Europe for an autonomous vehicle and this environment will demand significantly more of an AV developer’s AI.

Not what you were looking for?