Manufacturing operations will change dramatically in the next few years. Research by market intelligence and advisory firm IDC projects that by 2021, 75 percent of enterprise applications will use artificial intelligence (AI). Discrete manufacturing, along with retail and banking, is leading the charge to adopt AI for predictions, recommendations, enhanced customer service, and intelligent process automation.

An area in which AI can make one of the most significant impacts on manufacturing operations is machine vision. Analyzing production images can enable a smart manufacturing facility to:

• Immediately identify misassembled or flawed products, helping to achieve low PPM defective rates
• Automate aspects of quality control, lowering labor costs
• Minimize losses due to defects and waste
• Detect anomalies in production which could cause equipment damage and downtime
• Eliminate the need for workers to enter areas where they could be exposed to harmful substances
• Read barcodes and identify products or packages to ensure accurate shipments and no misplaced inventory

“Seeing” Has Been a Challenge for Manufacturers

Although analyzing production images can be extremely beneficial to manufacturers, there have been limited options for putting eyes on production lines and products. Typical vision systems now use cameras connected to a frame grabber module in a host computer. Vision software then processes images from the production line.

A substantial downside to this type of system is the high volume of data it produces, requiring a VPU/GPU integrated vision solution or a new heterogeneous computing platform. It’s often a challenge for manufacturers to optimize these types of vision systems, attempting to balance memory requirements, speed, power consumption, performance, and reliability.
 

ADLINK & Arrow Electronics Delivers a Next-Gen Vision Solution 

ADLINK and Arrow Electronics have partnered to provide a better machine vision option for manufacturers. Rugged ADLINK hardware performs reliably, even in the harshest environments, and provides native support for AI solutions. This enables us to collaborate with Arrow to provide total machine vision edge computing solutions, well-suited to manufacturing use cases.

Our solutions that bring AI machine vision to the edge overcome many challenges that manufacturers now face. Our AI at the edge solutions use embedded computers that enable data extraction and computation right at the source. A machine vision-based system can use crucial image data, such as a production defect, broken equipment, or mislabeled packages, and share that information where it’s needed most for crucial decisions or actions. It can be deployed to stop operations to correct faulty assembly or to alert technicians that vital assets or equipment need repairs before further damage occurs. Additionally, AI machine vision at the edge can confirm that products and packages are labeled correctly and are staged for proper shipping to satisfied customers.

AI at the edge will also help you scale efficiently as your smart factory operation expands. As more devices, components, items, and assets are connected to your system, your system may experience bottlenecks or delays. Edge systems enable computing at the data source, so latency, created when data first transmits to the cloud or a data center, is eliminated. Data can freely flow to the people, processes, or systems that require it. Additionally, if a network connection is interrupted, edge computing ensures mission-critical operations can continue reliably.

An ADLINK-Arrow AI machine vision-based system also gives you the advantage of a plug-and-play solution that deploys quickly, has a long product life, and minimizes the investment you need to make to add machine vision functionality to your operational platform. Furthermore, by working with ADLINK and Arrow, you have the advantage of support from organizations that understand your organizational needs, such as satisfying requirements for both IT and OT teams and end-of-life product management.
 

Achieve New Milestones

With AI machine-vision based systems, you gain additional advantages: advanced functionality and the opportunity for innovation. ADLINK-Arrow next-generation vision systems accomplish the functions that camera and vision software systems perform. However, AI machine vision at the edge can also give your operation enhanced security, compliance, and faster insights at scale. You may also be able to collect more data through optical character recognition (OCR) or confirm through facial recognition technology that authorized personnel are operating equipment or performing maintenance.

What’s more, AI machine vision at the edge will take you one step closer to your vision of the ultra-efficient, automated smart factory that will help you maintain your competitive position in your market as the industry evolves through digital transformation.

For more information, download our solution brief on heterogeneous computing for AI at the Edge.