• According to two new investigations about artificial intelligence and IoT in industry, the market is expected to grow up to 90.6MM€ by the year 2026.

• Use of artificial intelligence increases productivity in industry up to 15%, according to KPMG.

• McKinsey points that 56% of interviewed people reports having adopted AI in at least one of their business units. A figure that is 6% greater than it was on 2020.

These are but a few examples of the impact that artificial intelligence is having in Industry.

When speaking about artificial intelligence, many people picture autonomous machines, smart robots or other disruptive solutions that will completely revolutionalize industry in a few years.

Others remain skeptical, which is not surprising. Tech-related media and technology provider companies like us freely employ terms such as artificial intelligence and machine learning. As a result, the expectations set for this kind of technology are quite wild in many cases.

That is why, we want to bring many of these promises down to earth, so that manufacturers can truly see what can be expected from a technology as fascinating as artificial intelligence is.

Predictive maintenance

Predictive maintenance uses algorithms to foresee the next failure of a component, machine or system. This way, operators can perform specific maintenance procedures to prevent said failure at the optimal moment.

One could say it is about solving little issues before they become huge problems.

This way, companies can reduce unplanned stops caused by equipment failure, and decrease the number of costly interventions. Besides, maintenance no longer operates on a cyclic calendar, but instead revolves around the needs of the machine, being able to plan optimal maintenance dynamically.

The best way to see this in practice os through the example of one of our customers. It is a company from the machine tool sector that, thanks to employing anomaly detection algorithms in real time, is now capable to produce warnings when a component begins to wear out or behaves in anomalous fashion. Like so, the maintenance team is able to schedule a stop and amend the problem adequately, preventing the potential breakdown of the component in question.

Smart recommendations

Artificial intelligence not only can help companies in anomaly detection or potential future failures, but can also provide suggestions for improvement. Deployment of algorithms enables producing insights and recommendations based on process data.

One clear example on how to apply this technology comes from one of our users from the food sector. This company has a long and complex production process, with a duration between 4-5 hours where multiple factors need to be considered. For instance, parameters such as temperature or humidity are critical to the manufacturing process, and an oversight in said parameters may cause the final product to be rejected.

In order to homogenize the process and guarantee quality in the final product, data has been collected from different machines, artificial vision cameras and sensors. Thanks to self-learning systems, an algorithm has been develop that allows making recommendations based on current process data and environmental factors. The system produces various recommendations depending on the different parameters that can be present in the plant. Thus, the company has managed to reduce the amount of rejections and achieved their manufacturing goals.

Process optimization

Optimization of processes can be one of the most complex tasks in any factory. Performance of a process depends on a huge number of factors, and therefore, advanced methods and tools are required to optimize and improve production. Through the use of tools based on artificial intelligence and machine learning, it becomes possible to better understand the workings of the process and thus improve it, in a seamless way.

One important success case in this context is about one of our customers from the automotive sector. By means of the integrated edge & cloud solution from Savvy, data has been collected at very high frequency from machining processes, to later analyze the most relevant variables and thus detect inefficiency points. This way, they were able to improve the cycle time of their machines up to 16%, which means they have a greater yearly output of parts with the same number of machines.

Quality assessment

Regardless of companies producing automobiles, parts, foods or beverages, quality remains one of the key performance metrics in industry.

An example of artificial intelligence focused on quality can be observed in one of our customers, that had a persistent problem in their quality tests that made them reject actually good parts. Savvy has carried out a descriptive analysis of all the tests performed for various months with the aim of obtaining trends about test results and spot false positives. Thanks to this analysis, an algorithm has been elaborated to warn about the reliability index of any test performed. This algorithm has been deployed on the Edge, running in real time and reporting results to the cloud simultaneously. Consequently, the company has reduced the number of wrongly discarded parts, thus reducing costly waste.

Conclusion

Artificial intelligence can help industry in many ways, not only in the areas covered above, but also in many other fields such as supply chain management, warehouse management, process automation, cybersecurity, etc.

The most important aspect is to identify different points in your plant where AI can provide the most value. It is not simply a miracle fix, but applying it in the right fashion and with sense, it can bring huge benefits that directly translate to monetary profit.

If you believe our success cases align with the challenges you face in your day-to-day, do not hesitate to contact us.