MyAdvantech Registration

MyAdvantech is a personalized portal for Advantech customers. By becoming an Advantech member, you can receive latest product news, webinar invitations and special eStore offers.

Sign up today to get 24/7 quick access to your account information.

Advancing Highway Safety in Japan with AI-Powered Traffic Surveillance


AI plays a crucial role in enhancing highway safety, mainly through advanced traffic monitoring and management systems. These AIpowered systems offer real-time analysis and can swiftly identify potential hazards, including accidents, cargo truck spills, and traffic violations. With capabilities in event detection and data collection, these systems provide a comprehensive security solution for enhancing surveillance accuracy. Additionally, AI contributes significantly to active traffic management by facilitating dynamic adjustments based on traffic conditions. These lead to reduced emissions, resource conservation, and prompt recognition of safety-critical events. In an effort to take a more proactive stance on highway safety, a leading Japanese industrial solution provider offered the latest AI-powered solution to upgrade the highway traffic surveillance and monitoring system.


The project involved several challenges. First, the constrained space inside facility cabinets, coupled with the need for uninterrupted 24/7/365 operation, necessitated a compact and reliable edge AI box. Furthermore, situating these boxes roadside, especially considering Japan’s temperature extremes and limited ventilation, posed substantial design challenges for the electronic systems. Second, budget constraints faced by the customer, the complexity of protocols, the expansive geographical distribution of the system, and a shortage of field service personnel all added layers of difficulty to ensuring regular system maintenance. Third, the nature of motorways demanded that the edge AI boxes possess substantial AI inference processing capabilities and several AI models to handle the multitude of objects, including jaywalking pedestrians, vehicle recognition, adjustment for changes in weather and environmental conditions, and potential hazards. Finally, the complexity of the system required a solution that could accommodate multiple high-resolution cameras covering various angles at intersections or across lanes with high-speed traffic.multiple high-resolution cameras covering various angles at intersections or across lanes with high-speed traffic.


To address these challenges, Advantech Japan proposed the AIR-030 Compact and High-Performance Edge AI Box, which comes with longevity support until 2030. Complying with industrial standards and boasting a wide operating temperature range, the AIR-030 not only features a fanless design to mitigate heat dissipation commonly associated with high-performance computing solutions, but it has also successfully passed strict reliability and durability testing for CB and UL safety certifications.

Utilizing the NVIDIA® Jetson™ AGX Orin 64GB module, the AIR-030 offers impressive performance, boasting up to 275 trillion operations per second (TOPS) across multiple concurrent AI inference pipelines. It is compatible with NVIDIA JetPack™ SDK 5.0 and above. Leveraging resources from the NVIDIA® Jetson Ecosystem, the highway traffic surveillance solution provider greatly reduced the development time for multiple AI models, successfully addressing the increasing complexity of traffic scenarios.

The AIR-030 is also equipped with three 2.5GbE Ethernet ports, providing stable high-speed data transmission to the traffic control center and enabling the connection of up to five megapixel cameras for vision AI processing. The AIR-030 comes pre-installed with the necessary OS and drivers verified by Advantech, such as Ubuntu Linux 20.04 OS with long-term support (LTS). Advantech’s commitment also extends to after-sales service, maintenance updates, and ongoing technical support. This ensures that the AIR-030 remains current and secure, providing users with a reliable and forward-looking solution.

The integration of AI into the AIR-030 has successfully eliminated reliance on cloud services for video analytics, resulting in significant cost reductions. This achievement is attributed to the AIR-030's high-efficiency computing capabilities, allowing for AI inferencing to be directly performed on real-time video data captured at the edge. By transmitting only the analysis results to the cloud, the new system significantly minimizes data transfer and cloud storage expenses. Furthermore, with the AIR-030 integrated into the traffic control center’s traffic content management system, real-time traffic event notifications can be displayed on roadside screens. This helps drivers respond immediately to unexpected changes in traffic conditions, enhancing overall safety on the road.


Advantech Japan provided timely technica l consultations on the development tools and OS throughout the project development. They also offered in-depth knowledge on the heat dissipation design of the facility cabinet, including enhanced vents for airflow optimization. This support allowed the solution provider to focus on AI model development and deployment, thereby significantly reducing both development time and costs.

The completed AI-powered highway traffic surveillance and monitoring system can now detect numerous situations and safety threats, such as cargo spills, smoke, vehicles travelling in the wrong direction, stopped vehicles, or queues. By delivering real-time 24/7 surveillance and intelligent analysis, enhancing the traffic control center’s decision-making capabilities, enabling drivers to respond appropriately to emerging conditions, the proposed solution has ultimately helped improve highway safety in Japan.


• Superior robust design with fanless heat dissipation

• Timely local technical consultations and longevity support until 2030

• High-efficiency computing capabilities for AI inferencing on real-time video data captured at the edge


• Real-time 24/7 traffic surveillance and intelligent analysis

• Detection of numerous situations and safety hazards, such as cargo spills, smoke, vehicles travelling in the wrong direction, stopped vehicles, or queues