Areg Romanowski

The virtual proving ground environment to test the AD algorithms is implemented by the famous 3D game engine UNITY. The vehicle behavior is simulated by the integration of FMU 2.0 model that can be generated by tools such as OpenModelica, MATLAB ... The data processing algorithms can be implemented by C/C++ or python running in parallel by the UNITY engine where the data is transferred by shared memory. Fast and real-time. See the details in our white paper which you can download from the link. The company introduction can be downloaded from this link.

• 1-677-124-44227
• 437 S Olive St, Los Angeles
• Mon - Sat 8.00 - 18.00h

We provide a sample proving ground to demonstrate the notion of the concept. Here the proving ground is a 2 lane road with the necessary curvature to test the LDW. The road has a damaged texture and the lanes are low-quality faint. The road has different luminosities to test and validate the robustness of the tested algorithms.  The sensor set is consisting only of a front camera with a typical aperture and resolution. You can download the environment from the link. The executable can be executed from windows and the prerequisites are defined in the read me files.

Maybe the most well know and common Level-1 feature used in today's vehicles is Lane Departure Warning. The feature utilizes the front camera and detects the lanes. If any unintended lane departure is detected the feature issues a warning. The necessary Visual C++ project and source codes can be dorwnloaded from the link. This is a standalone executable that runs in parallel with the UNITY environment which processes front camera video using the well-known image processing library OpenCV.