How AI is impacting the manufacturing sector
By Arvind Saraf
There is an old saying that goes, “You can’t improve what you can’t measure. This is why AI is having such an impact on data-rich industries. But should AI be used to measure machines or humans? Let’s take a close look at an industry, manufacturing, where machines have been largely measured for years, while humans have not been effectively measured.
While there is a lot of talk about how artificial intelligence is transforming the way manufacturers source data and analytics to understand and solve some of the biggest problems they’ve faced when it comes to machines, very few much has been discussed over a large data pool that has been inaccessible for over a century – actions performed by humans.
Despite all the hype surrounding automation and robotics in manufacturing, the simple fact is that humans still do most of the work on assembly lines. A closer look at the numbers makes it clear that it will be at least 100 years before automation is more abundant than humans in manufacturing. A recent survey shows that up to 72 percent of manufacturing tasks are still performed by humans and that 71 percent of the value created by the operation comes from human actions. This figure is not surprising when you consider the cognitive ability, adaptability, and dexterity that humans bring to the table.
However, human activity on the production lines has effectively remained invisible to analysis. Trying to harness this invaluable data, let alone put it to good use, has long been a challenge. Unlocking that data becomes all the more important when you realize that nearly 340 million humans (according to a study by Goldman Sachs) are still employed in various factories. These data further strengthen the argument that human performance analysis is a clear requirement for manufacturing success for the foreseeable future.
The limits of time and motion studies
Conventional methods for monitoring human production using stopwatches or physical observations are fraught with several limitations. This process is time consuming, labor intensive, susceptible to observational bias, does not capture behavioral data accurately, and can lead to incomplete documentation and analysis. Also, these processes can only measure the time it takes to complete the process and fail to observe other variables that need to be measured, such as process stickiness and the quality of the output. Such limitations make it difficult for companies to make accurate assessments of the performance and effectiveness of their employees and the processes followed – and therefore to make sound decisions about personnel, production, training and more.
The complexity of human actions adds to the challenge. This is especially true for manufacturing. Although it seems like a simple task, assembling a printed circuit board, for example, involves several steps, which makes the process vulnerable to variability. The aforementioned survey points out that 73 percent of variability in factories comes from human workers, not machines. While a robot can be programmed to do a job a thousand times without variation, human output will vary in quality, duration and more. The result may also differ from person to person, as no two people are the same.
AI provides comprehensive data – and insight – on human actions
So how can manufacturers solve this problem? As it is, video is the best way to capture this information in raw format. And thanks to breakthrough innovations in advanced data analysis techniques, such as computer vision and machine learning, manufacturers now have the ability to bring together these real-time footage, along with the associated data, that can be shared at all levels, from line workers to plant managers to improve efficiency. This not only had a huge impact on process optimization, but opened up new avenues for acquiring meaningful data and information about manual assembly.
AI at work in the factory
Standardized labor is the gold standard in manufacturing: it’s how to carry out a process that maximizes efficiency and quality. Many manufacturers struggle to ensure that standardized work is tracked on every station, every shift. This is where AI plays a key role. Advanced computer vision algorithms can be used to validate and measure critical elements of manual production lines using proprietary action recognition technology. This technology learns standardized work instructions from a line and observes workers as they complete steps, identifying any deviations in the manufacturing process, such as missing a key step or an out-of-sequence action by their workers. In some cases, the line associate is notified immediately; in others, the video is used for training purposes. This helps to minimize production defects.
Here’s another example where AI can observe humans for the benefit of manufacturers. Every discrete assembly line experiences bottlenecks, where one station is slower than the others and reduces productivity down the line. Today, manufacturers use time and motion studies or simple observation to determine the cause of bottlenecks, but it’s very imprecise. For example, on an assembly line it might appear that station three is the bottleneck, but a shortage of materials at station two can be the real culprit, which is not entirely obvious in one. a look. However, a consolidated view of objective data for the entire line makes these hidden causes much easier to discern. With this powerful information, manufacturers will have the ability to make important decisions such as providing additional training for the line worker at station two or improving the flow of materials, to reduce the downstream delay. AI enables continuous automated measurement of this data over time, allowing all areas of operator-specific improvement to be tracked.
So to answer the question – “What should AI, humans or machines measure?” The two are not mutually exclusive. It is important for manufacturers to have a clear picture of what is going on with the machines in their factory, but it is also essential to address this human-shaped blind spot that has been on the assembly line for more than a year. century. And to do it right, AI is a great tool because of its ability to process every action, turn it into data, and present clear analysis and information to manufacturers when and where they need it most.
If you have an interesting article / experience / case study to share, please contact us at [email protected]