How can artificial intelligence transform manufacturing?

By: Sophie Weaver

14, May, 2019


Artificial Intelligence -

patricio davalos 1201994 unsplash

Industry 4.0, have you heard about it yet? It’s a major buzzword these days and Artificial Intelligence (AI) is the fundamental component of this new revolution that the manufacturing sector is witnessing. Industry 4.0 enhances the computer systems and automation technologies -introduced during the Industry 3.0 phase- with the help of smart and autonomous solutions that leverage new-age technologies like machine learning, AI, IoT and Big Data.

How crucial is AI for manufacturing?

The answer is VERY.

“Why?” you ask. Its scope of possible applications, ranging from virtual designing of solutions and real-time maintenance of systems to creation of a smart supply chain and development of new business models, makes it a go-to solution for improving the overall production process.

A recent survey on artificial intelligence by Forbes Insights unveiled that 44% of all the respondents from the manufacturing and automotive sectors consider AI as ‘highly important’ to the manufacturing function in the upcoming five years, while 49% deemed it as ‘absolutely critical to success.’ [1]

AI leads the way for Smart Manufacturing and Maintenance

The application of AI for enhanced and efficient production and continuous maintenance is gaining traction as manufacturers realise the importance of AI in effective and timely identification and maintenance of the production line and the technology’s role in reducing the downtime. A major share of expenses is accounted for by the continuous maintenance of production line equipment. This can weigh down heavily on the bottom line of asset-oriented production operation. Reports also indicate that manufacturers record annual unplanned downtime costs of USD 50 billion, with 42% of the unplanned downtime resulting from asset failure.

Predictive maintenance

Both unplanned and planned downtimes can be easily tackled with the help of predictive maintenance, facilitated by AI. Predictive maintenance leverages AI’s key components- artificial neural networks and machine learning to formulate predictions related to asset malfunction. The AI algorithms continuously assess the production process and flags anomalies (even microscopic faults) in order to take timely maintenance and repair actions for reducing the risk of unplanned downtime. It can also help in prolonging the Remaining Useful Life (RUL) production machineries. The minute-by-minute update on production gathered and studied by the AI algorithms offer the technicians with insight into the components that need inspection in the cases where maintenance is necessary, and is also capable of suggesting suitable tools and methodologies. This can facilitate pre-scheduling of maintenance tasks and focused repairs.

Quality of production

With customers expecting faultless products and companies are under constant pressure of very short time-to-market deadlines, offering a product with high quality that meets quality standards and regulations can give a manufacturer sleepless nights. Industry 4.0 has given way to the new Quality 4.0 model, under which traditional quality methods are combined with the new-age technologies like machine learning, big data and AI among others to attain operational excellence. AI algorithms can alert the production teams about emerging faults like subtle abnormalities in the functioning of machines, deviations from recipes and other such issues that can cause issues with product quality.

Generative Design Software

AI is also capable of improving the way we design the products, as it can recommend solutions based on the detailed brief provided by theengineers and designers by way of generative design software (AI-powered software). The brief can include different parameters like available manufacturing methods and material types as well as restrictions like time and budget constraints. The software can process the detailed brief, explore different permutations and combinations and recommend most optimum product options. The recommended solutions can then be tested for actual performance across different manufacturing conditions and scenarios for finding the best of all proposed solutions.

AI’s role in optimising supply chain

In addition to product designing and manufacturing process, AI can also enhance manufacturing supply chains. The AI algorithms can study consumer behaviour, weather, socioeconomic and macroeconomic, political status and geographic patterns to formulate estimations for market demands. This information can help producers in predicting changes in the market and optimise energy consumption, inventory levels, supply of raw materials and staffing to better respond to the market changes.

Some examples of AI in manufacturing

British company Rolls Royce is not only a luxury automaker but also provides maintenance services for the airlines that utilise its products. Owing to this, the company, which works with a lot of aircraft, receives a lot of data. The company has deployed Microsoft Cortana Intelligence Suite and Azure to analyse this chunk of data. The solutions can perform data modelling at scale and identify and flag anomalies. The process helps the clients to plan ahead.

Korean company LG CNS is renowned for its cloud-based smart factory service, which helps manufacturers to automate their manufacturing process and monitor the production efficiency simply at the hit of a button. The company has opted for Microsoft Azure DocumentDB and HDInsight to gather production histories and uses Azure Machine Learning that can project emerging defects and ensure efficient production and management.

Automated machine learning software solutions provider Fero Labs offers a technology that is used by several steel makers to reduce ‘mill scaling’ that causes loss of 3% steel. The AI solution helps in reducing mill scaling by 15% and save millions of dollars [2].

German company Siemens has retrofitted its gas turbines with hundreds of sensors that transmit data to an AI-enabled data processing solution, which adjusts fuel valves to maintain the minimum possible emission level.

Future of AI in manufacturing

AI can help in improving manufacturing efficiency at reduced costs and improve the sustainability quotient of the production processes. This has led several companies into assimilating AI in their production cycles, with the aforementioned companies just a handful of examples. AI in manufacturing is here to stay as the AI in manufacturing market is anticipated to surpass USD 16 billion by 2025, up from USD 1 billion in 2018, as per a 2019 Global Market Insights report [3]. The technology, which is yet in its early stages, is being developed and explored further to find newer benefits and potential applications across different sectors.








Other sources: