AI in the factory


It is very common to notice how much technologies change products that people use in everyday life. At the same time, less attention is given to the revolution in the way companies manufacture products. And in the meantime, Internet of Things, Big Data and Artificial Intelligence have found their place in factories, about what Amit Chattopadhyay, the Director of Big Data and Industrial IoT at Western Digital (WD) told us.

In pursuit of the best products – how Western Digital is changing its factories

Today almost all factories are automated at the local stage and well adapted to each production line. As Amit Chattopadhyay described, in the case of hard drives, one device spends about 1 sec at each test station. By the time it reaches that station, several thousand device features have been measured. The data is Big as well as Fast. At current state of automation and test architecture, there is no path to utilize all of this thousands of variables to make an ‘intelligent’ call about the device. As a result, much of the Big Data remains Dark and is eventually discarded.

As Amit Chattopadhyay said, one can describe changes that are happening in factories by using an example of small-scale prototyping. When the scale is small, employees work on one thing at a time. They use their problem solving capabilities, data assimilation power and cognition in a real time to produce a robust high-quality product.

A Cognitive factory mimics the way humans learn and adapt, but operate in a mass production environment. In practical terms, this implies creating an intelligent automation system that can utilize the thousands of features in the device and provide a much-improved recommendation in less than a second. Development projects of this kind that employ machine learning to improve production processes are generally at the early stage of the journey; however the prospects of Artificial Intelligence in this context are quite promising. In addition to expectations about increasing quality of products, Western Digital aims at reducing testing time what will provide new information and in effect will allow making better business decisions. When algorithm provides human with conclusions drawn from pure data, it will be much easier to make right decisions.

Another important goal is to increase reliability. It will enable factories to manage unpredicted occurrences during the manufacturing process. It is a particularly important issue, especially when taking into account the fact that production process of hard drives has several thousand steps. Making the process more intelligent and cognitive will allow detecting problems ahead of time. That takes on particular significance in the light of the fact that factories belonging to companies are not new, but they have to be more and more efficient. For this reason, among others, Western Digital aims at using Artificial Intelligence in some fashion in all factories, applying machine learning on a large scale.

Potential benefits of machine learning applied to the manufacturing process, and creating cognitive factory are huge not only in the hard drive market. It is an opportunity for any sector of the economy, where physical product is an outcome of the process. Widespread application of Artificial Intelligence in factories will have a considerable impact on maintaining competitive advantage because – to mention only these most important – it will decrease costs and waste, increase reliability and efficiency, as well as contribute to significant improvement of quality. And all thanks to the use of machines in the areas where human abilities are not enough.