Entrance

Thanks to the continuous development of AI Edge Computing, many new AI systems are being developed to provide solutions to a wide range of applications. An application that is very interested in the use of AI Edge is the use of Smart Parking systems in garages or indoors.

Smart Parking works by using License Plate Recognition technology to check-in and check-out when cars enter and exit. Generally, the system also determines the time it takes to park the vehicle and automatically calculates the parking fee.

Smart Parking systems were initially developed using artificial vision technology. However, it has some limitations in machine vision-based systems. To recognize the license plate, the car needs to be steered at the right point and angle, which can cause frustration for drivers. Artificial vision systems perform poorly in dim light conditions or weather conditions such as fog or rain. Artificial vision systems are also likely to have problems with reading on special or non-standard plates.

To overcome the limitations of artificial vision systems, a company has developed an AI Edge-based solution for a large retail chain. They turned to AAEON to design a hardware platform that could power their systems to the highest potential.

Challenges

The developer faced several challenges in developing the AI Edge Smart Parking solution. The system had to work in Edge with little reliance on the cloud, work outdoors in an embedded environment, and also provide support for multiple cameras and sensors.

Edge System

To provide fast service to customers, the AI system must be able to operate at the endpoint. Relying on a cloud-based system can slow down the service as the system communicates with the central server. Parking lots and garages often have inconsistent network connections that interfere with cloud-based systems.

Outdoor Conditions

The system must be placed in car parks or garages which have customers. A compact system can provide flexibility, but most importantly, it is experienced in the point that the system can operate under a wide range of conditions and is not affected by dust and exhaust.

I/O Support

The AI app requires multiple camera inputs to read license plates when cars are in and out. Additional connections with controllers such as parking doors and payment kiosks help improve the usefulness and effectiveness of the system.

Solution

Working closely with the developer, AAEON has determined that the award-winning BOXER-8120AI AI Edge fanless embedded system is the best solution.

NVIDIA Edge Solution

The BOXER-8120AI family of AI Edge systems features powerful NVIDIA Jetson TX2. BOXER-8120AI delivers the performance needed to run AI on Edge with Hexacore SoC, NVIDIA Pascal 256 CUDA core GPU, 8 GB of memory and 32 GB of eMMC storage.

Embedded Design

Produced at just 30mm thick, the BOXER-8120AI is built to fit wherever it is needed. Working with fanless design is not affected by dust or exhaust. BOXER-8120AI is designed to operate at cold temperatures from -20°C to 50°C (-4°F to 122°F).

Designed for Modularity

The BOXER-8120AI has four Gigabit Ethernet ports, perfect for connecting to LAN cameras. It also has additional I/O ports, including HDMI, which provides security personnel with another tool to help keep shoppers safe.

Effect

With the BOXER-8120AI, the developer introduced an AI Edge license plate recognition solution to provide power for customers' smart parking spaces. Thanks to the power of the BOXER-8120AI, the developer's AI system provided functionality beyond machine vision-based solutions. The system was able to recognize and read plates from different angles in different lighting conditions and in fog, rain or other harsh weather conditions.

When a smarter smart parking solution was implemented, company owners were happy to have a system that was easier and faster to park for their customers, and did not involve problems caused by machine vision-based parking systems. Supported by AI Edge license plate recognition and the award-winning BOXER-8120AI, Smart Park helps improve the experience of shoppers and store owners.

Product

The award-winning BOXER-8120AI AI Edge embedded fanless system is part of AAEON's family of BOXER-8100AI embedded AI solutions. The BOXER-8120AI features 256 CUDA cores and a range of AI frameworks including Tensorflow,Caffe2 and Mxnet, as well as the hexacore NVIDIA Jetson TX2, which supports user-developed AI inference software.

The BOXER-8120AI has four Intel LAN (i211) ports to support IP cameras and also features a range of robust design features, including dust prevention features, a wide range of operating temperatures, a wide range of power inputs, and an aluminum chassis. BOXER-8120AI’s compact, cost-effective design and powerful processing capabilities mean you can deploy your AI platform anywhere.

AAEON Manufacturer Services and AAEON Turkey Official Distributor ICC provide configuration, customization and OEM/ODM support to ensure fast and efficient delivery of your AI application. Whether it's replacing an I/O port or restructuring the BIOS, ICC and AAEON are a stable hardware solutions partner and ensure that your project meets your specific needs.