Italian electronics manufacturer Egicon has been using emerging data analysis tools to fully automate production since 2017. In the process, it reduced the repair rate by 80%, eliminated scrap, improved warranty support, and shortened the delivery time for real-time quality reports from one month.

Headquartered in Modena, Italy, Egicon produces electronic control units, instrument panels and human-machine interfaces for the automotive, agricultural, biomedical and aerospace fields.

Egicon integrates Siemens' Valor and Opcenter Execution Electronics IoT software into its production and quality systems to achieve continuous monitoring and provide customers with better warranty support and traceability data.

“We were able to reduce the repair rate from 30 parts per million to 6 parts per million and achieved a scrap rate of 0% in 2019,” said Michele Magri, production manager of Egicon, in a published case study. "Now I can get instant updates of all manufacturing processes without leaving my desk. I can spend my time on innovation and improvement."

Other software manufacturers also reported significant results.

Cobus van Heerden, senior product manager for analytics and machine learning software, GE Digital’s Proficy software has helped manufacturers in many industries gain many benefits, including a 90% reduction in waste, a savings of $5 million in quality improvements, and an 80% reduction in For GE Digital, the downtime said. A company gained key insights on how to control its dehydration chemicals to obtain the best quality within a few hours.

According to a published case study, FactoryTalk Innovation Suite is a joint product of Rockwell Automation and PTC, helping Rockwell achieve a 33% increase in labor efficiency, a 70% increase in output, and a 50% reduction in training time.

These and other emerging data analysis tools are overcoming the limitations and obstacles of their predecessors.

According to Izik Avidan, business unit manager of Siemens Digital Industry Software's digital manufacturing analysis business, a major obstacle in the past was that the tools that provided potential advantages have not yet been used.

He said that more than 80% of advanced analytics projects have failed, and this statement is supported by research by Gartner and other institutions.

"From a manufacturer's perspective, the main problem with data analysis tools in the past is that they are still tools," Avidan said. "Many platform and solution providers don’t realize that ordinary manufacturers don’t have all the necessary skills to fully utilize these tools. The tool provides the functionality it was designed for, but the entire project may fail. You must be able to speak in the manufacturing language and all of these Build bridges between new technologies. Manufacturing customers do not have these skills."

"Historically, you really need a PhD in mathematics or data science to get value from analysis," van Heerden said. "You need to put the analysis in the hands of their existing operations staff. You can't say to manufacturing customers, "You need to retrain your employees or hire new employees to benefit from the analysis." The key is to let the process engineers and the production line operate Members can access the analysis."

Ed Cuoco, PTC's vice president of strategy and solutions, said: "Tools are designed for experts to make it easier for experts to deal with difficult challenges, not to simplify their work." Many tools also require on-site data scientists. The end result: "These tools are not suitable for large manufacturers, they usually don't have their own data scientists," Cuoco said.

According to Avidan, Cuoco, and van Heerden, other obstacles include:

Failed to understand and solve the pain points of the manufacturer.

Manufacturers are required to replace expensive old equipment tools.

The inability to access the necessary data to gain insights is usually because the data is in an isolated system, sometimes called dark data. Between 60% (Forrester) and 97% (Gartner) collected data is still unused.

Data that cannot be easily combined with other data.

Data that is difficult to clean, format, and prepare.

Assuming that the data meets high-quality benchmarks, in many cases experts are needed to improve data quality.

There is a lack of analytical tools that enable managers to take action.

An analysis tool that is too difficult for ordinary operators to use.

Tools that cannot be expanded beyond the initial pilot or demonstration.

Today's tools provide fast value realization, simpler operation and scalability. Avidan said that more and more manufacturing software vendors understand that their customers need platforms that can combine multiple tools and integrate well on the shop floor.

 "We are now seeing more success with analytical tools in overcoming these obstacles," van Heerden said. "The tools we provide are showing evidence of rapid value."

Cuoco said that software manufacturers are designing tools and platforms that will run in factories that use 40- and 2-year-old machines. "These tools need to work in a real-world environment," he said. "This is the key to the applicability of the factory. This allows the factory to function within its capabilities without requiring them to be good at areas that do not belong to them."

 Emerging tools provide the ability to access, store, and process data. Subject matter experts are available onsite or remotely. The cost of ownership is low. It does not require too many additional servers or cloud resources. This is easy. Avidan said, configurable, customizable and capable Provide some value immediately.

"Nowadays, most software companies understand that putting some machine learning solutions on the shop floor will not solve your quality problems," he said. "Now, they provide a complete turnkey solution, which may change the rules of the game."

Avidan added: "My professional life is under such tension, trying to provide out-of-the-box solutions, while also understanding the need for customized solutions, flexible customized solutions to optimize the needs of manufacturers." "In the past. In the past five years, we have seen more and more hybrid projects, both platforms and tools, combined with software tailored specifically for the type of industry."

Cuoco said the industry has not yet reached the point where tools are available out of the box, similar to the iPhone.

"Out of the box is the direction," he said. "Out of the box is the goal. Our solutions are beginning to become mature enough to see this."

Avidan says these tools are sometimes combined with experienced manufacturing engineers to help manufacturers improve performance and predictive maintenance, and integrate quality control into production.

Van Heerden said that tools are also "surpassing alarms", allowing humans to take action to become more closed-loop-the tools themselves can take real-time and safe control actions to enable factories to obtain or maintain optimized productivity.

Cuoco said that more manufacturing software vendors no longer provide analysis only as a component of the product, but provide analysis in solutions that solve specific use cases.

"If you introduce the data patterns in the data set or the list of conclusions he has seen before to an experienced manufacturing engineer, he will be able to easily transform that data into actions that an operator, line manager, or plant owner can take and noticeably Improve results," Avidan said. "With this turnkey solution, we can solve most challenges within days or even hours."

Cuoco said that it is better to add more domain expertise to these tools. "There is a need to embed more and more domain-specific knowledge," he said. "How can we put experts and machines together so that they both understand the problem? The machine must be able to say,'I can consider parameters specific to the field.'"

Van Heerden said the analysis is also improving because they apply to the entire supply chain from raw material suppliers to shippers, manufacturers to end customers.

Cuoco said that improvements are still needed to make the tool easier to build, and the design of the tool allows the machine to do more work.

Avidan said that costs will definitely need to be further reduced so that small and medium manufacturers facing similar problems can use the technology.

He said that more standards are needed so that manufacturers can more easily integrate technologies from different vendors.

"We have to understand that there is already a lot of software in the workshop," Avidan said. "Any solution you want to introduce into this ecosystem must be well and seamlessly integrated into these IT solutions... In this way, if you have an action item that needs to be pushed from one engineering system to another, it can Do it in a single portfolio. This is one of the most important things you can do."

van Heerden said that as standards and interfaces become more open, integration will become easier.

Van Heerden said that to be successful, please work with trusted industrial suppliers to provide a comprehensive range of products.

"Many analytics vendors provide solutions that can solve part of the problem. Some people can analyze the data," he said. "Some people can make predictions. Some can run simulations. Another can optimize settings. Work with trusted suppliers who will not disappear tomorrow and can provide all these features in a single product."

Manufacturers seeking instant perfection and full integration should lower their expectations and adopt a step-by-step approach. "Until all systems are perfect, don't take the'big bang' approach to adopting technology," he said. "I recommend a quick, incremental approach. Equip operations staff with easy-to-use tools so that they can quickly gain incremental value."