5 ways you can use Machine Learning in manufacturing

By: James Green
Tag(s): Data
Published: Oct 27, 2020
5 ways you can use Machine Learning in manufacturing

Artificial Intelligence (AI) used to be the stuff of science fiction. But over the past few years, a particular kind of AI, Machine Learning (ML), has become a key staple for business leaders looking to bring out untapped value in data they already collect. In fact, half of businesses in one study said they expect ML to be key to delivering competitive advantage and that it will determine their company’s future success.

Machine Learning works by using large volumes of data to train sophisticated algorithms. These can then be applied to new data to identify hidden patterns and predict future outcomes. 

In the manufacturing sector, ML can be applied across the supply chain to deliver real business benefits such as:

  1. Improving operational efficiency and lowering costs, by using ML to optimise the factory floor. Take the example of Google, which was able to reduce electricity use in its data centres by 40% by using custom ML to control the air conditioning in its server farms. The improvement was achieved even though Google had already spent a lot of time manually optimising its processes. Google is not alone: a quarter of early adopters of ML have been able to increase the efficiency of internal operations thanks to ML and more than 80% say it’s helping them drive down costs. 

  2. Reducing maintenance costs and improving reliability, by using ML to develop optimised maintenance schedules based on the way equipment is actually used. The same approach can be extended to customers to provide them with personalised maintenance schedules. For instance, a trucking company was able to save millions of dollars on parts each year by understanding when each truck would need servicing based on the actual routes and terrain it had driven.

  3. Reducing inventory levels and waste, by using ML to more accurately predict demand and optimise production schedules. One textile manufacturer was able to reduce stock levels by 30% as a result of using ML to predict customer demand more accurately, which let it move to just-in-time production. Overall, a quarter of early adopters of ML say it’s helped them gain a better understanding of customers and prospects.

  4. Improving quality control on the production line, by using ML to identify faulty products. Baby food manufacturer Kewpie is using ML – in the form of the Google Vision API – to pick out discoloured potato cubes that, while perfectly safe for babies to eat, may cause concern for parents. The ML system has replaced "human" inspection, which is tedious and stressful for production line workers while also more error-prone and more expensive.

  5. Improving the design of new products, by using ML to understand how products are actually used, how they actually perform, and what causes them to fail. Those insights can be fed back to design teams, while ML can also be used to predict the performance of proposed design changes. A third of early adopters of ML say they’ve enhanced their R&D capabilities thanks to ML.

What’s clear from these examples is that the payoffs can be significant and fast. These aren’t isolated examples. Research by Deloitte found that a typical ML project will deliver an ROI of between 2x and 5x in the first year. No wonder that another study by MIT Technology Review found that 60% of businesses are implementing an ML strategy and that a quarter of early adopters are devoting more than 15% of their budget to ML projects.

Yet, even though more than half of companies believe ML will determine their future success, ML can seem out of reach. It requires massive and flexible resources and deep but rare expertise. That’s why companies are looking to the cloud to deliver their ML projects — with more than 85% of ML workloads running in the cloud — and why cloud providers like Google Cloud is developing easy-to-use tools that give every company access to these powerful technologies. 

Working with our data analytics and AI team

Our Data, Analytics and AI practice brings together a highly committed team of experienced data scientists, mathematicians and engineers. We pride ourselves in collaborating with and empowering client teams to deliver leading-edge data analytics and machine learning solutions on the Google Cloud Platform.

We operate at the edge of modern data warehousing, machine learning and AI, regularly participating in Google Cloud alpha programs to trial new products and features and to future-proof our client solutions.

We have support from an in-house, award winning application development practice to deliver embedded analytics incorporating beautifully designed UIs. If you'd like to find out more about how we can help you build your own modern data and analytics platform, why not take a look at some of our customer success stories or talk to our data analytics team.

Machine learning at a glance with Google Cloud


Article updated October 2020
First published May 2018

< Back to resources

Think big. Start now.

We don’t believe in Innovation, we live it. Innovation combined with pragmatism is what runs through our veins. We ask ourselves the same question over and over again: Does it deliver value? And how quickly? Your big ambitions can start now.