Ways to bring higher-order automating intelligence to the production line
Factories are increasingly deploying artificial intelligence and analytics, with manufacturing companies turning to digital automation to create cheaper and faster production processes.
Many of these companies are just starting to find ways to deploy these new technologies. A recent article in Enterprise IoT Insights, described how “factories are bustling with innovation experiments, as industrialist are availed of ever-more sophisticated digital trends.”
In the article, “AI—On the line,” I offered some insights and KPMG viewpoints into efforts to bring higher-order automating intelligence to the production line. Some companies have already made strides in the digital transformation of their factories. But many elements around these processes still have to evolve before manufacturing can reap the full benefits of AI and analytics.
Here are some of the main points I made in the article.
AI can be applied from one end of the production line to the other. But where it can really make a difference—and be far more effective than humans—is in the detection of product defects. This is especially the case when it comes to tiny assemblies, such as microscopic integrated-circuit designs. AI can be used for live defect detection right in the production line, potentially replacing hundreds or even thousands of people doing quality-assurance inspections at the end of the run.
AI can also play a key role in increasing production efficiencies throughout the supply chain, so manufacturers can create a balance between forecasting production in advance to minimizing expenditure and inventory, while at the same time making sure they can fulfill customers’ orders.
Creating new assembly processes is another aspect of manufacturing that can benefit from AI. By using AI to envision entire production lines in a virtual environment, manufacturers can test literally thousands of scenarios to come up an optimal design.
Despite these benefits to manufacturers, the use of AI still needs some refinements in several key areas.
One such area is decision-making. To be fully effective, AI needs a variety of data to make decisions. In some factories, a vast amount of data can be extracted from the production floor, as a result of sensors installed on manufacturing machines. However, data capture still needs to evolve significantly to be effective. At this point, data from factory machinery can be rough or broken or in a form that can’t be used because it lacks context. Also, sensors in the same locations at the same factory under the same conditions often produce different data. That disparity makes extracting any meaning from the data a challenge, at best, and at worst, impossible.
Another area is the use of the Internet of Things in the digital transformation of factories. These transformations will require all parties to collaborate to reinvent the production process. All aspects of the manufacturing line need to be connected to be able to make effective business decisions. The capabilities to bring about such wide-ranging transformation aren’t available in a pre-packaged “off-the shelf” solution. This is less a consequence of the immaturity of the market and more a function of each company having their own multiplicity of industrial processes that will make digital systems integration and digital solution design new every time.
To learn more about how AI is changing factory manufacturing, you can read the entire Enterprise IoT Insights article here.