How industrial companies are catching up in the AI race
Opublikowano 16 Grudzień 2019 w AI
Artificial intelligence (AI) is achieving breakthroughs in the industrial sector. McKinsey recently forecasted that the global market for AI-based services, software and hardware will grow by up to 25% annually and will be worth around US$130 billion by 2025. But AI also represents a major challenge for industries in Europe, which are lagging behind the US and China. So, how can we start putting AI into action? Predictive maintenance is one area that demonstrates its advantages and potential.
AI in industrial environments
The use of AI-based technologies and robots in industrial environments is still in its infancy. However, more and more manufacturers are recognising its potential for increasing Overall Equipment Effectiveness (OEE), reducing costs and boosting productivity. Companies in other industries should perhaps follow the example of the healthcare and automotive sectors.
Tim Foreman, Omron’s Research and Development Manager, comments: "Manufacturing companies are rather conservative, because they often work with machines that have to run for 20 years or more. But that doesn't mean they have to be left behind when it comes to artificial intelligence. It's time to look more closely at the opportunities and possibilities of innovative technologies in the industrial environment."
AI for companies of all sizes
European countries need to pay more attention to automation and AI. Until recently, some small and medium-sized enterprises (SMEs) thought that AI was too complicated and expensive. There are also many complex and unique designs in plant construction and mechanical engineering. Learning experiences can’t therefore be easily transferred from other machines. The complexity of most systems is usually so high that it isn’t possible to describe the entire system quickly and simply.
Because SMEs are often closer to customers, they can be more agile, but need to be able to keep their production costs down so that they can compete with larger firms. One answer is to use robots and AI solutions such as open source software and applications that rely on machine learning. Large technology companies such as Google and Amazon are pioneers of these.
To make the best possible use of these new opportunities, companies need to work with large amounts of data and advanced algorithms, the two cornerstones of AI. Managers and employees therefore need training in this area. Robotics and automation providers are currently developing AI integrators that will help SMEs to make practical and efficient use of AI. Companies therefore need to focus on the opportunities and expand their competencies.
AI at the edge
AI-based technologies help to forge a new harmony between man and machine: ‘factory harmony’. But this is only possible when AI projects are strategically and comprehensively planned and implemented.
Omron's AI Controller is the world's first AI solution that operates ‘at the edge’, (with the hardware based on the Sysmac NY5 IPC and the NX7 CPU) recognising patterns based on process data collected directly on the production line. It’s integrated into our Sysmac platform – a complete solution for factory control, including motion and robotics, image processing and machine safety. This can be used in the machine directly, to prevent efficiency losses. Starting from the largest efficiency problems, the findings and optimisation gained can be scaled up to the whole production floor.
Our AI Controller is easier and faster to implement than other solutions and machine-level AI is ideal for predictive maintenance and machine control. It combines line control functions with real-time AI-based data processing. This enables companies to quickly identify and respond to unforeseen situations in real time, improving quality, maintenance and machine lifecycles, and then scaling this as required.
Edge v cloud computing
Cloud computing involves simple and uncomplicated access to data and systems. However, in industrial environments manufacturers need a way of looking inside the machine and real-time performance verification.
Edge computing improves control and security and limits resources such as hardware and algorithms. Sensors that collect the required information directly at the machine enable deeper and more up-to-date data analysis. Important information can be consolidated and compressed, further optimising oversight and transparency.
While the cloud is suitable for managing large amounts of data and long-term analyses, AI at the edge is essential for real-time applications. Production lines and machines can be monitored with real-time sensors, and the data quickly collected and checked for anomalies.
Conclusion
Technologies such as the AI Controller rely on AI in its simplest form, but this type of pattern recognition will continue to evolve. To see a more comprehensive implementation of AI-based solutions in the industrial environment, AI must become more user-friendly and companies must have more confidence in themselves.
In this context, business aspects are always more important than technological considerations. Business goals and future plans should determine the use of technology and not the other way around. Ultimately, AI in the factory can be much simpler and easier than you might expect!
