Explore applications of Edge AI at the factory

As our manufacturing advances into Industry 4.0 era, the machines used in manufacturing require increasingly more sophisticated control systems. It is not just digital transformation of manufacturing but improving the process by combining IT with the IoT, Edge AI, and 5G technologies to improve the process holistically and provide increased workplace safety. Edge Computing for AI in Smart Factory is important field that embodies these advances and hosts these new applications, such as Smart Processing, RTLS, Smart Predictive, and Smart Monitoring.

Smart Processing – Aside from the automation and robotics, sometimes we still need people on the production line for inspection and repairs. With the latest technologies, we can use AR/MR plus 5G technology enabled by edge computing for real-time remote training and guidance. This can greatly expand the capabilities of inexperienced field operators and improving the success rate of field repairs.

RTLS (Real Time Location System) – The real time location system uses 5G, IoT sensors, and RF tags for real time location audit of equipment, movement of materials, and control the operator to optimize the factory efficiency with lowest latency and create data and logs for tracking production metrics.

Smart Predictive (Quality/ Supply chain) – Big data and machine learning can be used to collect enough data for analyzing the processes to predict ways to get better quality and supply chain.

Smart Monitoring – Utilizing computer vision, real time monitoring and deep learning, AI can help to predict and prevent potential accident and injuries at the factory. The AI enhanced computer vision can also help to safeguard the premises as well as enforcing keep-out-zones without using physical barriers. This can open up the factory floor space to allow easier movement of materials through the factory to improve production efficiency.

For Edge computing AI in Smart Factory, we recommend:
In the factory site, huge amount of data needs to be processed with low latency, SCB-1932 and SCB-1937 are well suited for these requirements. SCB-1932, powered by dual Intel 3^rd Gen Xeon SP CPUs, can support up to 2x dual-width GPUs, along with plethora of Network Expansion Modules up to 100Gbps for maximum throughput. It is also at home doing inference workloads with ability to support multiple NVIDIA T4 GPUs, allowing you to scale the systems according to your requirement. It is also compatible with AEWIN Secure Boot module for hardening the firmware against tampering. SCB-1937 is the AMD EPYC 7003 series counterpart to the SCB-1932. It has the same feature set, allowing you to choose the best CPUs for your specific applications.

SCB-1932
SCB-1937

SCB-1830 is powered by a single Intel Xeon-D 2100 series CPU inside a compact 1U chassis. This chassis is able to support NVIDIA T4 GPUs for inference computing, along with a complete set of SFP and SFP+ ports for advanced networking. QAT is onboard for accelerating crypto algorithms. This compact system can fit into compact racks to allow ease of deployment into almost any scenarios.

SCB-1830MEC

For modest computing requirements on the factory floor, SCB-6012/6013 ruggedized system with wide temperature range support is the perfect solution. SCB-6990 fanless system is also a great candidate. These systems are designed with connectivity in mind, featuring various wired network options as well as wireless support. These fanless systems eliminate the final moving part inside a computing chassis to achieve even higher reliability.
SCB-60112A/B
SCB-6013
SCB-6990B

In the smart factory, cybersecurity and powerful servers are getting important. AEWIN has extensive knowledge in building secure and reliable systems trusted by some of the top ISVs and Network Security experts as the foundation of their solutions. AEWIN has a comprehensive portfolio featuring the full line of Intel processors. Explore how flexible design and customization options can meet your stringent requirements of your solutions.