As the use of artificial intelligence (AI) and machine learning (ML) increases, Edge Computing systems are demanding more performance and specific processing capabilities in industrial settings. This places intensifying demands on industrial computing systems that need to deliver time-critical decisions. The addition of video or other AI/ML algorithms means images and sensor input is processed at the Edge where its low latency operation is optimized for time-critical decisions.
Whether the machine is sorting products using vision systems, or processing sensor input, real-time responses are critical. Processing at the Edge means machines can be controlled locally in real-time and any alarms and exception data can be sent upstream to the Fog or Cloud layers.
Furthermore, ‘Smart’ factories also use Edge computing to analyze data locally and sending only summarized data to the cloud, saving bandwidth and reducing latency for fast, accurate, reliable communications around the factory or plant.
Edge computing in an industrial setting typically requires additional performance considerations, such as reliable, uninterrupted operation in harsh, demanding environments. They are often subject to extremes of temperature, shock and vibration, high noise and high EMI conditions.
Industrial IoT Hardware – Secure the IIoT
In addition to the challenging operating conditions, designers also need to be innovative as space can be restricted in crowded industrial settings. The rugged WINSYSTEMS SYS-427 embedded computer is specifically designed for operating in confined spaces and harsh conditions. It is based on the Intel Apollo Lake-I E3900 processor and has up to 8GB DDR3L ECC system memory and a CPU clock rate up to 2.0GHz. It is also TPM 2.0 Hardware Security enabled.
Vision systems are increasingly being introduced in industrial settings. For video-intensive applications in vision systems, industrial and building automation, the ITX-P-C444 is WINSYSTEMS’ industrial Pico-ITX single board computer (SBC) is based on NXP’s i.MX8M application processor. The Arm Cortex-M4 microcontroller is designed for real-time subsystems while there are also dual Ethernet, industrial I/O, and expansion options to meet present and future industrial Edge computing requirements.
Both products support Mini-PCIe expansion for adding communication or TPU/VPU accelerators for improved AI inference processing.
Industrial Partnership
The specialist nature of many processes means that Edge gateways are a vital component of an industrial network to accommodate the various use cases, protocols and formats used by different sensors and other node devices dotted around the network. The gateway is also expected to transmit/receive data converted from more than one protocol, while maintaining limits on bandwidth, energy and product service life.
When both the hardware and software are delivered in a solution that operates out of the box, the upgrade to the network can be seamless, increasing both efficiency levels and ROI.
A partnership between WINSYSTEMS and ClearBlade provides turnkey solutions for industrial computing operations, based on WINSYSTEMS’ ruggedized computing gateways running IIoT software by ClearBlade.
The turnkey embedded solutions can be easily integrated into existing systems. Development time can be significantly minimized by implementing pre-tested, pre-configured hardware and software solutions that customers can customize for their specific applications.
WINSYSTEMS’ modular approach to IIoT gateways allows system developers to meet the precise computing, energy and efficiency requirements for processing, video and analytics in an intelligent network that operates reliably in harsh conditions. It can also integrate various I/O data points according to the network’s specifications, e.g. vision systems, displays, monitoring and control, without minimal operator training.
ClearBlade and WINSYSTEMS worked with a mutual customer on x86 and Arm-based embedded systems for their rail safety application before partnering to deliver turnkey solutions. The partnership will accommodate all of today’s industrial Edge computing needs but will also future-proof systems to enable AI and machine learning systems.
WINSYSTEMS has nearly 40 years experience in designing and manufacturing embedded computing products. The SBCs, COM Express modules, I/O modules, and panel PCs for harsh environments are all made in the USA in facilities certified to ISO 9001:2015.