200 million cameras push intelligent machine vision to the edge

Machine vision (MV) uses technology that enables industrial machines to “see” and analyze tasks and make quick decisions based on what the system sees. MV is quickly becoming one of the most core technologies in automation.

Given that this technology is now being merged with machine learning (ML) to lead the transition to Industry 4.0, the possibilities are vast, especially at the edge. ABI Research predicts that by 2027, total camera system shipments will reach 197 million units and revenue will reach $35 billion.

"The shift from machines capable of automating simple tasks to autonomous machines capable of "seeing" long-term optimization factors will drive new levels of industrial innovation. This is ML known as MV (also known as for computer vision,” explained David Lobina, artificial intelligence and machine learning analyst at ABI Research.

He added that ML can extend machine vision well beyond visual inspection and quality control by augmenting classic machine vision algorithms with the scope and reach of neural network models, This is a classic of traditional computer vision.

Opportunity

Of all the trends in the artificial intelligence market, the computing edge has the most exciting applications and advantages – namely those where In devices belonging to embedded systems and the Internet of Things. Smart manufacturing is perhaps the most direct example, where smart cameras, embedded sensors, and powerful computers can bring ML analytics to every process step.

Intelligent machine vision is at work in factories, warehouses and shipping centers to help and assist human workers by handling more mundane tasks, allowing workers to use their expertise to focus on the important parts .

The market development in smart cities, smart healthcare and smart transportation has also matured, including ATOS (urban sector), Arcturus (healthcare sector) and Netradyne (transportation sector). major suppliers in the field.

As with other cases of edge ML applications, the best way for technology to advance is through a combination of hardware and software solutions and the adoption of information-rich data. Only through a comprehensive approach that brings all these factors together can fruitful results be achieved.

Suppliers realize they need to offer competitive products. In cases where sensitive or private data is involved, such as healthcare, the entire package should provide hardware (cameras, chips, etc.). ), software, and a great way to analyze data.

The “blanket” approach may not be the most common example on the market. Still, vendors must become increasingly aware of how their products integrate with other solutions, which often requires hardware-agnostic software and software-agnostic data analytics.

"This is a crucial point for smart cities, healthcare and transportation, especially regarding what machine vision can achieve in all these environments. For edge MV, software and hardware vendors, and service providers will start to expand their view of the industry,” Lobina concluded.

The above is the detailed content of 200 million cameras push intelligent machine vision to the edge. For more information, please follow other related articles on the PHP Chinese website!