Lidar Developer AEye & Seoul Robotics Develop a 300-Meter Range Perception Solution for Intelligent Traffic Systems

author: Eric Walz   

NASDAQ-listed AEye, Inc., a developer of high-performance LiDAR solutions for automotive applications and smart cities, has teamed up with 3D computer vision and software company Seoul Robotics on a long range, intelligent perception solution designed for Intelligent Traffic Systems (ITS) and smart city applications.

Seoul Robotics' technology combines computer vision with AI and deep learning to improve the perception capabilities of autonomous vehicles. The company developed its own proprietary software, which it says is compatible with nearly all LiDAR sensors on the market from different manufacturers to increase their accuracy in classifying and tracking objects. 

Lidar, which is an acronym for "light detecting and ranging", is a laser perception technology that's used to detect objects, including other vehicles, pedestrians, bicyclists and other objects in the environment around the vehicle. Lidar sensors emit pulses of laser light which bounce off solid objects. The reflected light is used to create a 3D rendering of the environment.

AEye's long-range lidar-based perception system even has the ability to detect an object as small as a brick at a distance of 160 meters. This extended range gives autonomous vehicle software more time to react to hazards.

The partnership will combine AEye's 4Sight lidar and Intelligent Detection and Ranging (iDAR) software with Seoul Robotics' 3D SENSR perception software. The result is a complete perception solution for long-range object detection and classification.  

AEye says its iDAR is the only software configurable LiDAR sensor with integrated deterministic artificial intelligence. iDAR captures more information with less data, enabling faster and more accurate perception. 

The SENSR computer vision software from Seoul Robotics is used to create more accurate models of the world, in real time. For Intelligent Traffic Systems (ITS), it provides best-in-class 3D perception for vehicles, cyclists, and pedestrians with a range of over 300 meters.  

Lidar sensors emit pulses of laser light that reflect off objects to create a 3D representation of the environment. (Photy: AEye)

Despite the company's name, Seoul Robotics does not actually make robots, rather it makes their perception system highly intelligent by identifying objects in the environment and tracking their movements. When this technology is combined with lidar, it can better make sense of what lidar sensors "see" with each scan. 

"Our solution has three key advantages,'' Seoul Robotics CEO HanBin Lee told Forbes in January "It is sensor agnostic, it has great accuracy and it has great efficiency."

Seoul Robotics is further improving the safety of autonomous vehicles by utilizing 5G infrastructure. Instead of placing sensors and computers directly on board vehicles, the hardware is placed on the surrounding environment such as buildings, light poles and other stationary infrastructure such as traffic lights. 

The network of sensors in the surrounding environment are controlled by a central processing unit, which communicates with all nearby vehicles simultaneously and in real-time, using vehicle-to-everything (V2X) communications technology. AEye's architecture enables Seoul Robotics to process advanced feedback loops and optimized scan patterns, according to William Muller, VP of Business Development at Seoul Robotics.  

The technology can be used to classify track vehicles near intersections, traffic circles, 4-way intersections and toll booths. But with the addition of Seoul Robotics' 3D SENSR perception software, its now able to monitor high-speed traffic on divided highways for the first time.

AEye's intelligent LiDAR is software-definable, enabling system integrators like Seoul Robotics to customize it to their needs. Uses for intelligent lidar include intersection management, Automatic Incident Detection (AID) on highways, even automated toll collection. 

Coupled with long-range capability of AEye's 4Sight M, the integrated AEye-Seoul Robotics solution will be able to achieve object detection and classification in adverse weather conditions and at ranges not currently available on the market, according to AEye.

Seoul Robotics said it trained the SENSR's AI engine on AEye's 4Sight M lidar, so the perception quality is maintained in both day and at night, even in adverse weather conditions. It even works in heavy rain and snow using weather-filtering algorithms developed by  Seoul Robotics.

Smart city and ITS integrators can benefit from reliable classification of motorists on the road, even if some of the vehicles or obstacles are partially obstructed, fast-moving, or clustered together. 

Cities using the technology to manage traffic will be able to distinguish weaving motorcyclists during peak congestion and even detect road debris and alert road maintenance vehicles as needed.

"With the integration of the 4Sight M and SENSR, infrastructure developers and city operators using this solution will be able to evaluate data at high-risk intersections – including lane usage by vehicle type, vehicle counts per lane, accident rates, and near-misses – in adverse weather conditions and at distances beyond 300 meters," said  said Akram Benmbarek, VP of Business Development and Strategic Initiatives at AEye. 

Deep learning models enable Seoul Robotics' 3D SENSR perception software to track more than 500 individual objects simultaneously with an accuracy of just 10 centimeters. 

"We're extremely impressed with the adaptability of AEye's 4Sight M sensor, it allows us to optimize output for different ITS use cases," said Muller.

Autonomous vehicle research VSI Labs has independently confirmed and published the AEye's 4Sight M's breakthrough range, resolution, and speed capabilities.

Research firm Markets and Markets predicts that the Intelligent Traffic Systems market will grow from $17.9 billion in 2020 to $36.5 billion by 2025, with another study projecting the total available market for LiDAR in smart infrastructure to increase nine-fold from $1.5 billion in 2025 to $14 billion in 2030.

AEye is among a handful of promising lidar startups, including Volvo-backed Luminar, Innoviz, and lidar pioneer Velodyne Lidar Inc., that are working on advanced lidar systems for automotive ADAS and autonomous vehicles. The company was founded in 2013 and headquartered in Dublin, California. 

In July, AEye announced a separate deal with Tier-1 supplier Continental AG to supply its long-range lidar perception sensor, which will be bundled into a full-stack autonomous driving solution developed by Continental. The Full-stack solution combines lidar, camera and radar data for safe navigation.

Continental made a strategic investment in AEye in Oct 2020. The two companies did not disclose the exact amount of the investment, but said they would work together to compete for contracts from automakers to equip future vehicles with lidar-based perception systems to support automated highway driving.

AEye and Seoul Robotics will demonstrate the smart perception solution during ITS World Congress, which takes place Oct 11-15, in Hamburg, Germany.

Eric Walz

Originally hailing from New Jersey, Eric is a automotive & technology reporter covering the high-tech industry here in Silicon Valley. He has over 15 years of automotive experience and a bachelors degree in computer science. These skills, combined with technical writing and news reporting, allows him to fully understand and identify new and innovative technologies in the auto industry and beyond. He has worked at Uber on self-driving cars and as a technical writer, helping people to understand and work with technology.