Verifying ADAS accurately

What makes the OxTS RT-Range the preferred reference tool for testing and verifying advanced driver assistance systems (ADAS)? The RT-Range is the most comprehensive test solution for the development of ADAS sensors giving a large number of highly accurate vehicle-to-vehicle and lane positioning measurements. Many leading vehicle manufacturers world-wide rely on the compact ADAS test solution to benchmark and improve radar, vision and lidar sensors used in active safety systems. It is unique in that it offers a wealth of invaluable features which provide engineers with a real advantage in their research into active vehicle safety.

 

 

Collision Mitigation and Lane Departure Warning systems need to be tested and verified against company standards as well as international vehicle safety legislations. The RT-Range far surpasses these requirements and offers a magnitude of additional features which make testing ADAS quick, reliable and repeatable.

 

With the RT-Range you can measure the position and orientation of several vehicles simultaneously, in real-time and with exceptional accuracy. Positional accuracy can be measured to 2cm and velocity measurements are accurate to 0.05 km/h. Up to four Target vehicles can be used allowing engineers to test complicated traffic scenarios. With our clever Target system inside a backpack, even tests with pedestrians and cyclists can be carried out.

 

 

We are constantly working to include more features in the RT-Range. In the past we have developed many unique features like the time-to-collision and time gap outputs, longitudinal offset for balloon car testing, 3D range calculations, triggered logging capabilities, car-to-pedestrian tracking, graphical presentation of the vehicle positions with Bird’s Eye View, or the compatibility with Anthony Best Dynamics (ABD) in-vehicle robots. By combining the ABD in-vehicle robots with the RT-Range it is easier to benchmark cars, quicker to obtain repeatable results and safer to carry out tests which could be dangerous for human test drivers. Using the RT-Range with ABD’s robots allows engineers to trigger tests from the RT-Range measurements. For example you can trigger a lead vehicle to cut in front of an approaching vehicle at a specific range.

 

 

 

Another big improvement in the new software is the representation of Target vehicles as polygons. The RT-Range computes the closest range from the Hunter vehicle to a polygon representation of the Target vehicle, automatically swapping points and even interpolating between polygon points. Polygons are more representative of real-world vehicles compared to single points. Especially more advanced tests, like blind spot detection or collision detection at junctions, are easier to perform when a polygon is used.

 

 

Other advantages of using polygons include the ability to have hidden target vehicles. The compact system can compute whether a second target is visible and whether the radar should be able to identify it. This is needed to benchmark radar performance when acquiring new targets. The RT-Range even measures the percentage of the target that is obscured or hidden behind another target.

 

 

When testing vision systems for lane departure warning (LDW) it is important to have a measure of the curvature of the lanes. Using new outputs from the RT-Range the curvature of each line is computed and output over the CAN bus. The output is available in real-time so that it can be captured and compared to real-time sensor data. Another key measurement for LDW systems is the angle of the vehicle within the lane. The new RT-Range software now computes the relative heading of the line markings compared to the heading of the vehicle.