A leading DMS company used the iEPF-9010S series via a workload consolidation approach for phenomenal upgrades to their AOI system.
A leading DMS company used the iEPF-9010S series via a workload consolidation approach for phenomenal upgrades to their AOI system. One iEPF-9010S can handle all AOI tasks for product quality inspection, replacing the traditional setting of using two or more IPC systems and improving overall image data transmission performance. Results show up to a hundred times improved data transfer rate for increased product inspection efficiency with fewer devices to manage, hence reduced system equipment footprint and reduced system integration complexity.
A leading DMS company wants to build AOI systems with machine vision inspection technology to improve product quality and customer satisfaction. This DMS company uses industrial control PCs based on Windows OS to perform long term machine automation. When importing product inspections, a Linux system with NVIDIA® GPU runs inspection templates for image analysis.
Since the two operating systems are different, the company used two IPC systems to perform product quality inspections for the entire AOI process. An industrial control PC runs on the Windows operating system to capture and process images from the PoE camera while controlling the robotic arm. The other GPU computing system runs under Linux operating system and works with NVIDIA® GPU to process AI inference and image analysis for product quality inspection. However, evaluation via two IPC systems results in high latency due to frame transmission over a physical LAN. Therefore, the AOI station becomes the bottleneck of the whole production process.
The data transmission speed between the two IPC systems greatly affects the efficiency of AOI inspection. To solve the problem of data transmission latency between the two systems, ASRock Industrial provided DMS company with an iEPF-9010S to perform all AOI tasks through Workload Consolidation approach as a tailor-made solution . The iEPF-9010S system runs on Ubuntu Linux operating system and configures two virtual machines with a KVM hypervisor. Virtual Machine 1 (VM1) runs on the Windows operating system to capture the camera image and perform motion control. In contrast, Virtual Machine 2 (VM2) runs on the Linux operating system to perform image analysis and AI inference. After the virtual machines are configured, transmitting data between the two virtual machines over the VLAN results in a 10% increase in transmission speed compared to the physical LAN between two systems. We further designed a software tool to use shared memory for both virtual machines. It then effectively extends the transmission speed limit, greatly increasing the transmission speed by 100 times compared to the data transmitted by physical LAN in the traditional framework.
Supercomputing power, rich I/O and extensions to power the AOI system
The iEPF-9010S series features a 12th Gen Intel CPU for powerful computing capability, rich I/O, and flexible expansions to connect multiple industrial devices and strengthen AOI systems. Motion control is enabled via Motion Card to connect the PLC and motors to move inspection items through the production line conveyor belt. Through PoE connected to the camera and lighting controllers, the AOI system takes photos of the inspection items for fault detection running on the powerful Intel processor with NVIDIA GPU card and displays the real-time display of the results product inspection via the VGA port. After receiving the result, the controlling robotic arm transports the item to the next station. For other equipment maintenance capabilities, the digital output ports connect the tower light to indicate any status/condition abnormalities to signal maintenance alerts.
Workload consolidation and shared memory show 100x improvement in data transmission
The iEPF-9010S runs on Ubuntu operating system and uses the KVM hypervisor to allocate CPU cores and hardware resources in the virtual machine. The first virtual machine (VM1), with Core 1-6 processor running on Windows operating system, works as an industrial controller, performs motion control and captures images via PoE camera. With Core 7-8 CPU running on Linux OS, the second virtual machine (VM2) performs AI inference and image analysis with NVIDIA® GPU card for quality inspection. Transmitting data between virtual machines over a virtual LAN can increase efficiency by 10% compared to using a physical LAN between two systems.
However, the 10% data transmission improvement is not enough to match the throughput of the other production module, so the AOI station remains the bottleneck of the whole production process. To solve this problem, we designed a custom software tool, taking advantage of KVM’s Ivshmem functionality, to use shared memory between two virtual machines. This tool provides seamless accessibility for both virtual machines. It greatly increases the transmission speed of image data up to one hundred times faster than the data transmitted through the physical LAN in the traditional client frame. Thanks to this solution, the AOI inspection station is no longer the bottleneck of the manufacturing process, providing greater customer satisfaction.
- 100X Data Transmission Enhancement: Through workload consolidation and shared memory approaches, the new AOI system powered by the iEPF-9010S series reduces data transmission time and improves speed by up to one hundred times faster than physical LAN transmission for upgrades in overall product inspection efficiency.
- Fewer devices to manage and less system integration complexity: one iEPF-9010S for AOI system replaces the traditional setting of using two IPCs, significantly reducing system, maintenance and labor costs ‘work.
- Reduced system equipment footprint: From two traditional IPCs to an iEPF-9010S, the single-device solution occupies the smallest possible factory floor space, maximizing production line layout and ensuring efficient use of space at the most optimal level.
Did you enjoy this great article?
Check out our free e-newsletters to read other great articles.