AI – Partnering Computing Power and People Power for Manufacturing Efficiency

AI – Partnering Computing Power and People Power for Manufacturing Efficiency

Manufacturing has always been about efficiency, from Henry Ford developing the assembly line to the introduction of machines to perform standardised jobs on the factory floor. Today, efficiency in manufacturing being transformed by artificial intelligence (AI) and it’s closely-related cousin, machine learning. While automation is a big part of this process, AI requires critical support from specialist teams with the right know-how, making the human touch an essential part of successful AI applications in manufacturing.

Data, Data, Everywhere

Thanks to today’s technology, data can be collected at every single point of the development, production, distribution, or product lifecycle that a company can imagine. Huge volumes of data are collected every day in most manufacturing facilities thanks to the powerful computers that have become key to modern manufacturing. But this data is meaningless and overwhelming if it isn’t crunched, analysed, and injected into a digestible form that we can use. And that’s where machine learning and AI come into play, turning pure data into contextual, relevant data that can be used to improve key aspects of production teams.

AI Adoption in Manufacturing

AI is slowly showing its value in the manufacturing sector, most notably in the following areas:

• Defect detection – Quality control is a time-and-resource-consuming, often using hard-coded algorithms to determine if an item is functional or defective. These systems cannot learn or anticipate defects that may occur – they can only detect what they are programmed to detect. Machine learning and AI allows systems to adapt to new information and detect a broader range of issues in real-time, reducing false positives and hours of quality control processes while keeping within compliance regulations, and supporting the work of compliance teams. A good example of this is Audi’s AI and machine learning system, which utilises high-resolution cameras and image recognition software for deep learning, supplying design teams and manufacturing teams with more of the right information, when they need it most.

• Integration on the assembly line – Huge amounts of assembly line data are routinely collected, but there’s no ability for one set of data to interact with data gathered from other points in the assembly line. As a result, we can’t see the relationships between the different data or extract its full meaning and potential. AI and machine learning allow this data to fully integrate, pulling from the IoT of the assembly line to create a big picture view for management and manufacturing teams to effectively utilise in both day-to-day operations and larger strategic processes.
• Real-time troubleshooting – With thin margins, it’s more important than ever to tackle problems quickly as they arise, and powerful AI technology helps engineers to do this. Using sensors, IoT equipment, and data analysis, manufacturing facilities can troubleshoot everything from bottlenecks in production and scrap rates to equipment wear, all as it is occurring on the assembly line. This is also essential as a component of predictive maintenance, and a good example of this is the Thales SA tool used for high-speed rail lines.
• Demand forecasting – One of the most important aspects of AI and machine learning is the ability to forecast with greater accuracy. This is especially important in the consumer packaged goods industry, where the system can improve planning coordination, marketing/sales, and account management, finance, and supply chain management. At the Danone Group, this process resulted in a 50% reduction in demand planners’ workload (freeing up resources), a 30% reduction in product obsolescence (reduced wastage), and a 20% reduction in forecasting errors.
• Research and development –  AI and machine learning can also be used to help product development teams to innovate and solve challenges, thereby improving customer experience, aligning products more closely with what customers are looking for, and reducing time spent on product development. This process is currently being used by Nissan to develop new vehicles, and another Danone Group initiative is focussed on using AI and machine learning to improve products, teaming up with Brightseed to use AI to help uncover the full health and nutritional benefits of foods.

AI and machine learning are behind some of the most exciting and innovative developments in the manufacturing industry and are set to become a more commonplace feature throughout the world as 5G is rolled out and the IoT grows. However, it cannot replace the human touch. Instead, it enhances the way we can deploy specialists and teams, allowing computers to take on time-and-resource-heavy tasks while supplying teams at every level with real-time, actionable information that allows their expertise to shine.

At Otto, we work to bring the human touch to the world of IT, allowing our clients to harness the power of effective new technologies while ensuring that a specialist is always on the other side of the line, ready to help. Chat to your managed IT services partner to find out more about introducing state-for-the-art technology to your business without sacrificing the value of human creativity, experience, and drive.