From Hindsight to Foresight: The Power of AI & Prescriptive Analytics in Manufacturing

In 2017,? Tesla’s Gigafactory in Nevada?came under the limelight for producing a lot of scrap. Internal documents reviewed by Business Insider showed that 40% of the raw materials had to be either reworked or scrapped at the factory. However, by 2019, the company reported that it had managed to improve its yields by 75%. ?

Cases of reactive problem-solving are common in manufacturing and they come with huge financial burdens. But is it better to optimize processes beforehand or after facing the consequences? ?

Enter prescriptive analytics. Unlike descriptive analytics, which describes “what happened" after an event, or predictive analytics, which forecasts "what might happen," prescriptive analytics goes a step further. It analyzes vast amounts of data to not only predict potential issues but also recommend specific actions to address them.

Thankfully, technology has advanced and manufacturers today can not only leverage prescriptive analytics but also power them with artificial intelligence (AI).

In the case of Tesla, for example, AI-powered prescriptive analytics would have allowed the company to rapidly optimize processes. By analyzing real-time production data, the system could have identified deviations from optimal settings and recommended what remedial actions to take for the most efficient assembly line.

Prescriptive Analytics is the Future

Prescriptive analytics gives sixth sense to manufacturers. It helps them play out what if scenarios with different factors and identify what process improvements could reap the most benefits whether in terms of cost savings or efficiency. ?

Thanks to AI and machine learning, prescriptive analytics is on its way to becoming the norm in the manufacturing industry. AI and machine learning algorithms excel at handling large datasets. They can sift through sensor readings and machine logs, process data, and identify subtle patterns and relationships. They can then make precise predictions about potential issues and generate nuanced recommendations for optimizing processes.

The best part about AI-powered prescriptive analytics is that they allow manufacturers to analyze data in real-time and make immediate recommendations and adjustments for the production floor. Moreover, AI algorithms refine their predictions and recommendations over time as they are exposed to new data and real-world outcomes.

Prescriptive Analytics Use Cases in Manufacturing

Here are a few use cases where AI-powered prescriptive analytics can add significant value:

Station Design

Imagine a scenario where circuit boards assembled at a specific station on the line have a higher failure rate. Prescriptive analytics will not only highlight the issue but also pinpoint its root cause and the remedy. For example, in this case, AI-powered prescriptive analytics could recommend that workers adjust their hand movements in a way that doesn’t cause stress on the board during assembly. ?

AI algorithms can analyze a workstation and give suggestions on how to improve the placements of certain objects or improve hand movement to increase motion economy. It can also show the impact of these adjustments both in terms of cost savings and efficiency.

Line Balancing

Let’s talk about line balancing. Descriptive analytics will point out that some stations have much longer cycle times than others. For example, Task A (adjusting gears) takes 4.8 minutes, which is significantly higher than the takt time of 4 minutes, and creates a bottleneck at that station.

One solution is to balance the line manually. Industrial engineers can try out multiple scenarios and see which one works the best to achieve takt time goal. Alternatively with AI, industrial engineers can work out all the possible scenarios within minutes and get suggestions on how to best balance lines along with their outcomes. AI algorithms not only allow users to easily adjust multiple factors such as the number of workers and takt time but they also recommend optimal factors for the most efficient assembly lines.

Safety

According to the National Safety Council, the total cost of work injuries in 2022 stood at $167.0 billion, including?wage and productivity losses?of $50.7 billion,?medical expenses?of $37.6 billion, and?administrative expenses?of $54.4 billion.

Given the high costs associated with injuries, it only makes sense for manufacturers to be proactive about shop floor safety, and the best way to do that is to ensure factory floors and workstations are designed with ergonomics and worker well-being in mind. AI-powered systems can analyze video footage of workers performing tasks and identify potentially risky postures, repetitive motions, or awkward movements that could lead to musculoskeletal disorders (MSD). AI algorithms can then predict which workers might be at higher risk for developing MSDs and personalize prescriptive recommendations to prevent injuries before they occur.

Embrace the Future with Prescriptive Analytics

The integration of prescriptive analytics powered by AI and machine learning is revolutionizing the manufacturing sector. By shifting from reactive problem-solving to proactive prevention, manufacturers can not only avoid costly production inefficiencies but also enhance worker well-being and safety.

Whether it's through improving station design, balancing assembly lines, or enhancing workplace safety, AI-powered prescriptive analytics offers significant value and time-savings for industrial engineers. The capability to analyze data in real-time and provide actionable insights allows for immediate adjustments on the production floor, optimizing processes and mitigating potential issues before they arise. As these technologies continue to advance, their adoption in manufacturing will undoubtedly become essential, driving the industry toward a future of foresight and innovation.

?