TECHNOLOGY

The goal of AI-SEE is to extend the Operational Design Domain (ODD) of automated vehicles to all weather and visibility conditions experienced 365 days of the year.

Taking technologies to the next level

To take automotive perception systems beyond the state-of-the art AI-SEE will combine complex hardware and software development.

The concept is built on the four main blocks as shown in the figure below.

  • A high resolution adaptive all-weather sensor suite with novel sensors
  • An AI platform for predictive detection of prevailing environmental conditions including signal enhancement and sensor adaption
  • A novel simulation path which allows to realistically simulate adverse weather near the sensor to adapt and test the system on both real and artificially generated road scenes
  • High definition maps with dynamic layers adaptable to changing weather conditions

All weather multi-sensor perception system supported by AI

The innovation

The project will deliver the first 24/365 high-resolution adaptive all weather multi-sensor suite building on an innovative novel AI perception-processing scheme for low visibility conditions.

 

Sensor-near simulation models for all active sensors for artificial generation of synthetic inclement weather datasets will be developed. This is expected to revolutionise simulation, with conversion of good weather neural network datasets into inclement weather datasets, thereby saving large amounts of money and time in testing and validating inclement weather sensor performance.

The all weather multi-sensor suite will include the development of: 

Source: OQmented

a gated SWIR-camera that will enable a post-processing pixel-level depth estimation

Source: Ibeo

a short-wave infrared (SWIR) Lidar with a novel SPAD receiver architecture

Source: Veoneer

a PolCAM- active polarimetric imager with congruent LIDAR data

Source: Bosch

a high resolution 4D MIMO Radar prototype with a dense point cloud 

Source: Ibeo

a high definition dynamic map to suppport environment perception