Challenge

With modern manufacturing processes generating an increasingly vast quantity of real-time data, engineers at Pirelli sought to both develop the organization’s data competencies and acquire a tool that would enable them to deploy a single set of statistical best practices across the enterprise. “We envisioned a way of performing analysis in a standardized software, generating reports in the same way [across the enterprise], and doing analyses in the best, most efficient manner,” explains Process Quality Engineer Sara Sorrentino.

Solution

Pirelli leadership greenlit a proposal to deploy JMP statistical discovery software as the enterprise standard for data management, access, analysis, and reporting. Automation and scripting features, as well as Python integration capabilities, have enabled advocates to develop sophisticated custom dashboards that have significantly consolidated the company’s analytics workflow. Outreach and in-house training have quickly brought domain experts on board. 

Results

“Our workflow is now very, very lean,” says Manufacturing Quality Engineer Massimo Pampurini who, with Sorrentino, led efforts to develop analytics capacity. “We have a consistent system that allows us to pull data directly from the database, create tables, perform queries, clean and filter the data, and perform our analyses, all internal to the software.” Regarding the potential of this system to transfer real value to the company through data-driven improvements, Sorrentino adds, “It’s a very powerful company impact — the impact is huge.”

Founded in Milan in 1872, multinational tire manufacturer Pirelli is today not only a leading producer of tires for the global consumer market; it is also a powerhouse of elite motorsport. The company’s rich sporting history, in fact, goes back more than half a century, during which time Pirelli is the exclusive supplier for the FIM World Superbike Championship, World Rally Championship, GT World Challenge, and — of course — the pinnacle of global motorsport, Formula 1.

The substantial innovations Pirelli has introduced as the exclusive tire partner to the FIA Formula 1 World Championship — a position the company has held since 2011 — present microcosmic insight into Pirelli’s dedication to performance. The five-compound range of slick and wet-weather tires Pirelli developed for the 2022 F1 season alone, for example, required thousands of hours in indoor testing and simulation, the development of tens of virtual and physical prototypes to be validated by thousands of kilometers of testing by nearly every team on the grid.

Performance and quality testing are a cornerstone of the Pirelli brand, a philosophy that carries through from R&D to commercialization and manufacturing at scale. An early adopter of industrial statistics, Pirelli embraced the Industry 4.0 movement at its inception with an enterprise effort to increase speed and efficiencies, partly through implementing IoT. Ever since, the company has collected, maintained, and utilized extensive data on all its products and systems. As digital technology has advanced, those systems are now increasingly integrated, and Pirelli has elevated change agents from within their ranks who advance new solutions for utilizing data analytics.

Process Quality Engineer Sara Sorrentino and Manufacturing Quality Engineer Massimo Pampurini are two such advocates. Based at Pirelli’s headquarters in Milan and both working in semi-finishing, Sorrentino and Pampurini are joining forces to expand analytics capability at Pirelli sites around the world by developing automation, consolidating and standardizing data workflows, and educating colleagues in modern industrial statistics.

“Nowadays all of our processes generate a huge amount of real-time data. We need to have the competencies — and also the tools — to analyze this quantity of data appropriately,” Sorrentino says. “We envision a way of performing analysis in a standardized software, generating reports in the same way [across the enterprise], and doing analyses in the best, most efficient manner. And we are pushing as much as possible to spread this knowledge [to Pirelli] worldwide.”

A single software solution for the entire analytics workflow

In the early days of digital transformation, data access and integration at Pirelli were managed purely in Excel, a tool that ultimately lacked sufficient capacity and sophistication to keep pace with the company’s needs. Even after Pirelli transitioned its system to a dedicated data collection application developed in-house, data importation procedures still moved relatively slowly, Pampurini says. Connecting the database to software that could solve both data integration and analysis needs was therefore a transformative innovation.

“We had heard many times about the power of big data analytics, but we needed the right tools to implement analytics successfully,” Sorrentino explains. That’s why Pirelli ultimately landed on JMP statistical discovery software as its standard tool for the analytics workflow.

“Of course, there were other alternatives, like R or Minitab, that may be powerful from a statistical point of view. But with JMP, we have that plus scripting, automatic data importation, integration, and generation of reports and journals. And we have the ability to create dashboards in JMP for data evaluation,” Sorrentino says. There is a high value placed on the consolidation of statistical power, customization, user-friendliness, and diverse functionality, which allows multiple steps in any data workflow to be consolidated into a single software solution, she adds.

JMP use at Pirelli had begun nearly a decade before, with individual engineers using the tool for basic data management and exploration. In the proceeding years, however, as Pirelli’s data needs grew in tandem with increasingly integrated production systems, Pampurini realized JMP had far more to offer than just a platform for analysis.

“Once I became aware of the full power of JMP, my whole mindset changed,” he recalls. “Connecting the software to a database, for example, we unlock big possibilities. And we’re now able to see our data from so many different points of view; JMP allows us to retrieve more information — and more valuable information.” Consolidating all the steps of a traditional data workflow into a single platform, he adds, epitomizes the principles of Lean manufacturing.

“Our workflow is now very, very lean. We have a consistent system that allows us to pull data directly from the database, create tables, perform queries, clean and filter the data, and perform our analyses, all internal to the software. Furthermore, we now have a higher focus: Whereas it used to be that we could get lost in the exploration, now the workflow is driven toward what we want to find.”

“We now have a higher focus: Whereas it used to be that we could get lost in the exploration, now the workflow is driven toward what we want to find.”

Massimo Pampurini, Manufacturing Quality Engineer

 

Data visualization, Sorrentino explains, is a critical part of any data workflow, and one that she and Pampurini have worked hard to formalize in the statistical best practices they are deploying to colleagues. Fortunately, this is another area where JMP excels, she says. “Looking at plots immediately gives us an idea of how a process is behaving. For example, we can use the screening platform in JMP to explore outliers.”

“This first layer of exploration is critical because it enables us to understand whether we’re going down the correct path, and if not, change our approach,” adds Pampurini. “The JMP Graph Builder is a very powerful first-layer tool, not just because it’s ideal for exploration, but because you can see a lot of variables together at once. This isn’t possible in other software.”

Dynamic, automated interfaces offer a 360-degree view of systems

The automation of data integration processes and the creation of dashboards that update in real time have only further optimized the utilization of data at Pirelli. Pampurini explains how he has deployed JMP scripts to further streamline data workflows, even making use of the tool’s integration with Python to transfer the team’s existing Python data-retrieval code. “The integration of JMP with Python is really interesting because you have the option to integrate codes you’ve already structured in Python,” he adds.

Dashboards now quickly provide real-time insight into machine performance and the quality level of materials. The primary outputs of this kind of analysis, he explains, are equipment performance and product-quality insights that will help the team to identify strategic areas for improvement. The secondary output is technical information on how different parts of the enterprise are functioning that can be fed back to the R&D and processing departments.

“The ability to gather this information very quickly means we can have a 360-degree view of our systems and improvement opportunities,” Pampurini says.

“We envision a way of performing analysis in a standardized software, generating reports in the same way, and doing analyses in the best, most efficient manner. And we are pushing as much as possible to spread this knowledge to Pirelli worldwide.”

Sara Sorrentino, Process Quality Engineer

“It’s very powerful from a quality point of view to understand the action plan and which improvements are needed,” Sorrentino adds. “These analyses never stand alone; rather, they bring us to an understanding of the actions that need to be taken to improve our process. We can understand, for example, which points in a process influence waste or scrap.”

The potential to transfer real value to the company through data-driven improvements, she adds, is “a very powerful company impact — the impact is huge.”

In one example, a single JMP script enables users to look at specific segments of their operations. An initial data-cleaning step is performed automatically simply by running the data download script, and a second round of data cleaning happens when users perform the selection. From there, dashboard buttons allow users to analyze data by timeline or drill down further into different layers of exploration and analysis. “The final output, in this case, is a bar chart with the capability indexes by quarter, by factory, and so on. And after calculating and extracting different tables, there is also a join operation by which you can access all the information in a single table,” Pampurini explains.

Developing analytics capability among domain experts across the enterprise

While Pirelli’s global headquarters has now standardized around JMP, Pampurini and Sorrentino are quickly expanding efforts to transition sites worldwide to the tool. “Changing the mindset is fundamental, as is training and providing support with a general set of guidelines for how to get started,” Sorrentino explains. As a result, the two have invested their time heavily in developing a single set of best practices and guidelines for the enterprise to accompany the rollout of JMP and JMP dashboards.

An internal academy they organize offers training developed with input from people inside Pirelli in accordance with company priorities. Internal projects are proposed, and any group within Pirelli that thinks they could benefit from developing a project in JMP can participate. Training has also been organized at other sites – like Brazil, for example, where many employees have asked for JMP licenses.

“We are making an effort to spread the knowledge of JMP to make people aware that by using this tool, they can do the same things they’re currently doing in other software, but do it faster and more easily,” Pampurini adds, and Sorrentino agrees: “We have to make people aware of how much time they can save with this software.”

The results illustrated in this article are specific to the particular situations, business models, data input and computing environments described herein. Each SAS customer’s experience is unique, based on business and technical variables, and all statements must be considered nontypical. Actual savings, results and performance characteristics will vary depending on individual customer configurations and conditions. SAS does not guarantee or represent that every customer will achieve similar results. The only warranties for SAS products and services are those that are set forth in the express warranty statements in the written agreement for such products and services. Nothing herein should be construed as constituting an additional warranty. Customers have shared their successes with SAS as part of an agreed-upon contractual exchange or project success summarization following a successful implementation of SAS software.