BETTER AUTONOMOUS EMERGENCY BRAKING BEHAVIOR THROUGH SENSOR FUSION

Sensor fusion for AEB in NCAP scenarios

With BASELABS Create Embedded and developers we can address both prototyping and series development of sensor fusion software. Data from all automotive sensors like radar, camera and lidar are combined using object fusion, grid fusion and lane fusion algorithms. The resulting unified environment model enables driving functions like AEB.

  • Driving functions like AEB become safer and more accurate through sensor fusion. More details on the benefits
  • Standardized software libraries with tooling support like BASELABS Create Embedded significantly reduce development efforts and shortening development times.
  • The NCAP test catalogue for AEB systems include different scenarios. Handling these scenarios requires the usage of sensor fusion for the environment perception of the vehicle. More details and free poster download

NCAP test scenarios

Loading the player ...

The NCAP test catalogue for Automated Emergency Brake (AEB) systems include different scenarios. Handling these scenarios requires the usage of sensor fusion for the environment perception of the vehicle. Sensor fusion increases the data quality of sensors and combines the strengths of multiple sensors. Due to the increased data quality, AEB systems are more precise, reliable and safe.

Order your poster: Euro NCAP test scenarios for AEB

With our info poster, you get an overview on relevant NCAP test scenarios which are defined for testing AEB systems. The scenario nomenclature consists of parameters for the scenario, the movement of the vehicles involved in the scenario, the vehicle types and their direction.

Moreover, the poster contains insights on how sensor fusion algorithms can be used to build state-of-the-art AEB systems.

Download your poster here.

To receive a printed version, please contact us.

 

 

Benefits

Less false alarms to prevent unintended braking

  • Sensor fusion allows to separate false-positive measurements from real objects.
  • For an AEB system this is an important feature in order to prevent unintended braking which might lead to a dangerous rear-end collision.

Faster confirmation time

  • In order to get enough confidence in an object hypothesis, a best practice is to confirm a single object over time by multiple successfully measurement associations.
  • For an AEB system, a higher update rate leads to smaller safety margins and thus an improved end-user experience.

Improved surveillance area for next generation AEB

  • Future evolutions of advanced driver assistance systems such as a city AEB for intersections and pedestrians need to observe more complex areas around the host vehicle.
  • This can be achieved with the combination of heterogeneous sensor technologies.
  • Sensor fusion is a methodology in order to address the challenges of different sensor input data representations and the object hand over task at the edges of the individual sensor field of views.

Suitable algorithms for series production

  • Safety critical software algorithms for automotive series products such as an AEB system have to be developed according to safety standards such as ISO 26262 for functional safety aspects and methods for ASIL B.

Please find more details on the benefits of sensor fusion for AEB on the poster.

From prototyping to series production

BASELABS Create Embedded provides relevant functionality for the implementation of the environmental model of an AEB system.

For automated vehicles that operate in urban areas and intersections, relevant objects may be those which cross the path of the vehicle. Detecting cross traffic participants properly is a challenge for environment perception systems, because e.g. front facing radar sensors cannot distinguish between a stationary object and a transversely driving vehicle within a single measurement cycle. Sensor fusion and tracking algorithms can overcome this limitation, This requires additional resources and enhanced modelling capabilities. BASELABS Create Embedded includes several sensor models for e.g. the MobilEye smart cameras that support the configuration and handling of cross traffic participants, e.g. to implement automated driving functions that fulfil Euro NCAP requirements.

The NCAP scenario catalogue also includes scenarios that require the proper recognition and tracking of vehicles in adjacent lanes. Sensors like radars typically perceive multiple reflections of close-by vehicles which is often denoted as extended object tracking. To reliably track these vehicles, sensor fusion systems need to be aware of this situation. BASELABS Create Embedded supports the configuration and handling of extended objects for radar sensors with built-in models, other sensors can be added by user defined models. 

A third aspect which arises from the NCAP scenario catalogue is the detection of stationary vehicles. The track proposer of BASELABS Create Embedded is capable of initiating track hypothesises from such objects.

Learn more about BASELABS Create Embedded.

Contact & further information

Top of page