Efforts toward social implementation of AI/IoT are accelerating toward the realization of Society 5.0. There are concerns about delays and information security when using cloud AI to process data acquired by IoT devices, so the use of edge AI using GPUs is essential. In retail, manufacturing, and smart cities, where demonstration experiments have been actively conducted, there are an increasing number of cases of transitioning to the actual operation phase accompanied by large-scale introduction. There, the issues in "operation and management" of edge AI devices installed in large numbers over a wide area became apparent.
Issues in “operation and management” of edge AI devices
On-site response in the event of a failure is costly and inefficient
When maintenance personnel go to the location where edge AI devices are installed and work, the work load increases in proportion to the installation area and the number of devices.
Depending on the content of the failure, even if the problem can be solved by restarting, it is necessary to check it at the site, which reduces work efficiency.
When updating firmware, it takes time to grasp the target device
It takes a lot of time to manually manage which areas devices are currently in use and are eligible for firmware updates.
It takes time and effort to update and adjust edge AI apps after app distribution
After delivering an Edge AI application, it may be necessary to make adjustments and updates to the application, for example, due to changes in the Edge AI camera position.
In order to implement post-distribution updates for a huge number of edge AI devices, it is necessary to visit each site and respond.
SyncLect Edge AI DMS is the solution
SyncLect Edge AI DMS is a high-performance and secure solution that centrally manages a large number of edge AI devices installed over a wide area through remote monitoring, reducing the burden on device management and improving efficiency.
Device life management for each installation location is aggregated in the cloud, enabling remote power ON/OFF
With edge AI, when a failure occurs, it is necessary to isolate the problem as to whether it is a hardware factor or a software factor. CPU utilization, memory utilization, disk utilization, CPU temperature, continuous operation time, and problem device data are continuously acquired and displayed on the web management screen to improve the speed of problem solving. This eliminates the need for maintenance personnel to visit the site to inspect the device.
Acquirable data
・Device ID ・Last update date ・Connection status ・CPU usage rate
・Memory usage ・Disk usage ・CPU temperature ・USB type
・Wi-Fi connection ・OS version ・Distributed version
・Security update ・Usage package (Jetpack, etc.)
Edge AI application distribution and OS/security updates can be performed remotely
In addition to updating the OS, security patches, and firmware in edge devices, it is possible to deploy edge AI applications to any edge device at once. In addition, data acquired by edge devices (AI apps) can be linked with applications such as Power BI that are easy for companies to use.
Edge AI apps can be calibrated even after deployment
After deployment, if the application needs fine-tuning, it can be done remotely on the provided UI.
Continuous and stable edge AI camera operation and accuracy quality can be homogenized.
Standard support for NVIDIA® Jetson™ products
Jetson products, including the latest NVIDIA® Jetson Orin™ series, are supported as standard, so developers with Jetson can use it with confidence.
Details of compatible products
・NVIDIA® Jetson Orin™ Nano
・NVIDIA® Jetson AGX Orin™
・NVIDIA® Jetson Orin™ NX
・NVIDIA Jetson AGX Xavier
・NVIDIA® Jetson Xavier™ NX
・NVIDIA® Jetson™ TX2
・NVIDIA® Jetson Nano™
Click here for inquiries
Macnica is collaborating with Headwaters Co., Ltd., which provides SyncLect Edge AI DMS, to solve the problems of large-scale operation when moving from demonstration experiments to the next phase, and to help customers considering actual operation of edge AI in projects. We will help you achieve success in the shortest possible time. If you are considering or are currently developing edge AI devices with Jetson and are interested in SyncLect Edge AI DMS, please contact us.
Click here for Jetson product details
