The current edge computing is mainly applied in industries or application scenarios with large data transmission, high security and privacy protection requirements, and real-time data processing needs, such as crowd detection and tracking, bright kitchen project, public security network monitoring, smart construction site safety management, and intelligent chemical safety production. However, in the face of the current situation of various industries in the Internet of Things, what kind of solutions can edge computing bring?
Current Situation:
Many devices are connected to the park, with a wide variety, making daily management and coordination difficult and complex.
How to operate efficiently in a digital way?
For group parks, multiple branch parks are usually managed, and there is a demand for localized management of user privacy data. How to achieve efficient unified management in the cloud while taking user privacy into account?
Edge Computing Solutions:
Provide low-latency, diversified interface, and local closed-loop management capabilities for the Internet of Things edge;
Local closed-loop management of privacy data, with all data collection, processing, and storage occurring at local nodes;
Non-privacy data is cleaned and summarized before being uploaded to the cloud for machine learning and training, continuously optimizing and updating local intelligent algorithms;
Diverse interfaces support the connection of various subsystems/devices in the park to the Internet of Things edge for intelligent management.
Current Situation:
No unified standard protocols, long integration cycles, and high costs;
Large data collection volume with high bandwidth costs for transmission;
Data collection gateways lack remote operation and maintenance capabilities, leading to high operational costs;
Data privacy and security protection, selective cloud upload.
Edge Computing Solutions:
Support typical industrial PLC and bus protocols for complex asset modeling;
Edge data processing (cleaning, computation, quality monitoring, etc.); local data routing, time-series storage, and openness.
Current Situation:
In the era of rapid development of autonomous driving, the efficient collaboration between smart transportation and vehicle networking is the cornerstone of safe autonomous driving. How can autonomous vehicles perceive complex road conditions in real-time? How can they quickly calculate corresponding countermeasures based on different scenarios? How can they achieve millisecond-level responses in emergency situations?
Edge Computing Solutions:
Smart transportation edge nodes analyze through video and radar fusion, and intelligent algorithms detect road traffic events (algorithm training in the cloud, execution at the edge), effectively providing collision warnings, traffic light phase push notifications, traffic flow awareness, and control, among other intelligent services in real-time.
Provide traffic managers with holistic monitoring capabilities and vehicles with high-precision positioning and mapping services, as well as traffic safety warning capabilities, enhancing driving safety.
Boshite Technology’s RK3588/3399 Edge Computing Box Empowers the AIOT Industry
Product Positioning:
Industrial-grade edge computing AI intelligent box——
Suitable for security monitoring, AI retail, autonomous driving, industrial inspection, AI artificial intelligence, medical and other intelligent terminals.
GPU Mali-G610
Network Architecture TensorFlow/MXNet/PyTorch
/Caffe
Solutions: Equipped with high-performance CPU RK3588 core board, independent NPU computing power chip RK1808*n
Product Positioning
RK3399 AI Edge Computing Box——
Suitable for security monitoring, AI retail, autonomous driving, industrial inspection, AI artificial intelligence, smart medical and other intelligent terminals
Product Parameter Configuration
CPU RK3399
GPU Mali-T864 GPU
NPU 4*RK1808
Computing Power Standard 12T (3.0T-24.0T optional)
Memory 4GDDR+1G*4 DDR+16/32/64 eMMC
Operating System Linux
Network Architecture caffe, TensorFlow 1.3.2, PyTorch, ONNX, Darknet
Algorithm Model INT8/INT16/FP16
Video Codec H265,H264,WMV,MPEG-1/2/4,VP8,VP9
Video Access 8 channels of 1080P
Solutions: Equipped with high-performance CPU RK3399 core board, independent NPU computing power chip RK1808*n
Boshite Technology’s self-developed edge computing boxes support secondary development. For more hardware solution needs, please feel free to contact Boshite Technology.
