Seoul Robotics , a 3D perception solutions company building a platform that uses deep learning AI, is partnering with ALP.Lab, an Austria-based test laboratory for automated driving systems, to provide 3D perception traffic safety systems for the DACH region (Germany (D), Austria (A), and Switzerland (CH)) and Central Europe. The LiDAR systems provide detailed insights into accident-prone intersections and are currently being deployed in Salzburg, Austria.

The European Union (EU) has set a goal of reducing overall road deaths to zero by 2050. Progress has been made across Europe; however, the number of bicycle crashes has remained consistently high with it being the only mode of transport that has not declined in fatalities over the past decade. Cyclists have had the greatest increase in serious crash-related injuries at 24%, with a high proportion of cyclist fatalities occurring at intersections.

In order to achieve the EU’s goal and save lives, Seoul Robotics and ALP.Lab developed a LiDAR-based system, powered by Seoul Robotics’ perception software, SENSR, to better understand interactions between vulnerable road users and vehicles.

Salzburg Research deployed the solution in cooperation with the Mobility Laboratory zukunftswege.at after identifying two high-risk intersections with multiple bicycle lanes in the city of Salzburg. The solution was designed to collect and analyze precise and high-frequency data of all road user movements and specific and complex interactions between vehicles and vulnerable road users.

Based on the comprehensive data gained, we can analyze road user behavior and conflict situations and derive valuable insights, which generally contribute to an increase in road safety for all road users and specifically enable improvement measures at the analyzed intersection.

—Siegfried Reich, Managing Director at Salzburg Research

Each Salzburg intersection has been equipped with four sensors and a LiDAR-Processing Unit running Seoul Robotics’ perception software. This software, SENSR, collects data from strategically positioned sensors for insights into the position and speed of objects within the intersection and is uniquely able to differentiate between cyclists, pedestrians, and vehicles in order to analyze their movements simultaneously and anonymously.

SENSR’s deep learning AI uses non-biometric data to accurately detect, track, and classify hundreds of street-level objects, even in harsh weather conditions. The 3D perception software is sensor agnostic and able to seamlessly fuse multiple LiDAR makes and models to get a full picture of the intersection and the connecting roads in one field of view. When deployed, the technology records high-quality data that will be utilized for unprecedented analytics.

Through this partnership with ALP.Lab, Seoul Robotics-powered solutions will soon expand across Austria, and into Germany, Switzerland, and the broader Central Europe region.