COMPANY CONFIDENTIAL. DO NOT DISCLOSE.
Seeing is believing
Johnson & Johnson Vision Care passes the benefits from streamlined manufacturing, improved process capability and quality along to the consumer, making vision care more affordable and accessible to all who need it. Engineers attribute manufacturing innovations to improved ability to act on the available data.
_____________________________________________________________________________________
Challenge: Though leadership recognized the importance of deepening analytics capability across Johnson & Johnson Vision Care, data access barriers and systemic inefficiencies prevented process engineering teams from acting on the available data. Change agents within the organization identified the need for a tool that would provide a more streamlined analytics workflow without incurring a significant training burden.
Solution: After evaluating a number of industry-standard tools, Johnson & Johnson Vision Care Process Engineering decided to license JMP®, adopting it as the process team’s standard data acquisition and analysis tool, as it offered a straightforward way to operationalize the process engineering workflow, moving from raw data to decision many times faster than before. JMP eliminates the need for engineers to work between multiple tools, explains Senior Process Engineer Brendan Malone. “One of the big drivers we had for moving to JMP was being able to put it all into one box,” he says.
Results: Best practices for data analysis have been standardized across the company through streamlined JMP workflows. Process engineers are now able to work more nimbly as JMP removes data access challenges, facilitates more efficient collaborations, improves business continuity, and speeds the pace at which Johnson & Johnson delivers products to its customers. Says Malone: “We went from needing four days to gather the data to actually having the data and a decision made in four hours.”
_____________________________________________________________________________________
Despite dramatic medical advances in the last half century, eye health continues to be undervalued and undertreated globally. The figures are staggering in scope: nearly half of the 2.2 billion people around the world with a vision impairment have not received adequate care to preserve or restore their sight. Humanitarian agencies estimate the widespread undertreatment of vision impairment is directly linked to cost and access barriers, particularly among marginalized communities.
Johnson & Johnson Vision is on a mission to change the trajectory of eye health worldwide by innovating solutions to these systemic hurdles. The company has, in recent years, introduced new smart manufacturing capabilities and streamlined supply chains to create cost-saving benefits that can be passed to consumers. At the same time, Johnson & Johnson has also identified just-in-time manufacturing as a strategy not only to drive down costs, but also to respond more nimbly to customer demands and, critically, deliver improvements to quality metrics.
The challenge of implementing these new strategies falls to Johnson & Johnson’s Global Process Engineering Group, which supports the production of billions of contact lenses every year at facilities in Jacksonville, FL, and Limerick, Ireland. While historically the purview of the process engineer was to fix mechanical issues, modern process engineers are tasked with using manufacturing data to improve processes and work leaner across statistical process control, data acquisition, problem solving and product disposition.
From reactive to proactive
Like many manufacturers today, Johnson & Johnson Vision Care has developed state-of-the-art systems architecture to support the vast number of data streams flowing from newly automated facilities. It is an extremely data-rich environment, and process engineers have had to adapt their methods accordingly. By 2018, explains Senior Process Engineer Brendan Malone, the
company’s Global Process Engineering Group was facing down the reality that existing approaches were unable to meet the demands of an intensifying data load. “We were recognizing that even though there was a significant amount of data available, we weren’t able to action that data quickly to minimize impact from manufacturing issues,” he explains.
At the time, the team’s systems, software and methods were fractured, and non-value-adding time required to go from raw data to information was a source of widespread frustration. In that frustration, however, process engineering leadership recognized an opportunity to consolidate tools and upskill Johnson & Johnson Vison Care’s entire process engineering organization. The goal was to grow beyond what was traditionally a reactive organization to become a proactive one. It was a message that resonated instantly with executive leadership.
“There was a drive from senior management to develop the capability of the entire Vision Care process engineering team. [Our leadership] believed in the power of data and wanted to drive the process engineering group to be more data capable,” Malone says. “But to [motivate] change, you really need to have something good you can give people to help them change.”
Overcoming data access challenges is the first step in a culture transformation
For Johnson & Johnson Vision Care, the single biggest obstacle to becoming a more data-capable organization was ready access to data. After years of collecting data from more than 40 manufacturing lines per site, the company’s data systems were heavily siloed from retrofitted equipment, incompatible databases and systems that saved duplicate records in multiple places.
Process engineers, Malone says, were spending an inordinate amount of time seeking access and querying disparate databases just to collect the data they needed to get started. An array of existing software licenses was another source of inefficiency, as the group had to jockey between incompatible formats and tools to complete a single workflow. IT teams were working diligently to transform the data to a more usable format, but process engineers had an immediate need for a single, scalable, flexible solution.
It was increasingly clear that the process engineering group would not be able to effect a genuine transformation in its analytics capability without first solving the access problem. But there was also a second imperative: The team needed a more sophisticated platform for statistical analysis that could be used by engineers at all experience levels, from novices using statistical applications for the very first time to Six Sigma Black Belts.
It was this second motivator that led process engineering leaders to JMP statistical discovery software – but the first that sealed the deal. “One of the big drivers we had for moving to JMP was being able to put it all into one box,” he explains; data access and all the analysis tools an engineer could need in a single software package. Better still, JMP could manage large data sets on a standard laptop and be used effectively out of the box without incurring a significant training burden. While not new to J&J Vision, several team members and managers had experience with the software, which helped instill confidence as the process engineering team adopted JMP for manufacturing support.
Workflows for quick data access and approachable analytics – with or without the coding
With its dynamic data visualization tools, including the interactive Graph Builder, JMP made it possible for engineers without a background in industrial statistics to pick up the software and get started right away – no coding necessary. “Data visualization is hugely important,” Malone says. “As process engineers, we use it to get an understanding of what’s going on with the product to our operations teams who manage the lines as well as the management teams who make business decisions.
“Nobody has yet convinced me there’s anything out there that can give me the visualizations I get in JMP as easily. That’s probably where the buy-in from the decision point came with the software: It’s a visual tool, so when you show people, they’re very, very quickly on board with it because they can see the capability.”
While JMP was appealing to those less experienced in statistics, Malone adds, it also offered a wide range of capabilities to those on the other end of the skill spectrum. With JMP, the company’s Six Sigma and data innovation experts could deploy machine learning applications, explore functional data, use text analytics and write custom scripts, among other things. JMP, he says, offered superior capabilities for statistical process control, automated reporting and deep dives into data analysis. In fact, the process engineering team’s data capability expanded in just two years from 3,000 rows of data to over 25 million.
Standardizing around a common tool motivates deeper adoption of data-driven decision making
There are benefits to building an economy of scale, and Malone says having a common tool allows Johnson & Johnson Vison Care to minimize the overhead of building out capability. In what he calls a “tiered response,” more statistically advanced individuals interested in acting as change agents within the organization have access to JMP Pro to solve some of the team’s most complex problems. Malone explains that tasks that once involved a large number of steps from ad hoc database query and migration between multiple software systems can now be done in a single step with the Query Builder tool in JMP.
One of the key benefits of this tiered approach is that advanced users can deploy customized workflows and dashboards that automate multistep processes to the broader community of JMP users. Scripted analyses help the whole group standardize around a set of best practices. “If we have a large group of people who are using JMP, we can then afford to have a small group of really highly skilled super users who can put time into doing the really complex tasks. And everyone reaps the benefits,” he explains.
The scalability of these benefits has been key to changing minds on a broad scale. Malone says it was through active demonstrations that he was able to win over the skeptics. “One of the most effective ways to get people who are embedded in [outdated approaches] away from it is to do demos with the actual data they're using and show them what they need.
“The real kick then is when you say, ‘Okay, you've done it once and you know you're going to need to do the exact same thing next week…. So just click that little green script button in JMP, and you've got your analysis done again.
“When people see that, they’re absolutely sold because they suddenly realize, ‘I only need to do this once, then all I have to do is update the data table.’ And they see you've just knocked three or four hours off the day’s work very quickly… so they say, ‘JMP is worth switching to because we get a long-term benefit.’”
Just eight or nine months after the JMP rollout, Malone says, the change was already palpable. “A lot of people [in other departments] were seeing what we were doing and saying, ‘How can we do that too?’ Now there is a huge interest in JMP… because [other groups] have seen the way we’ve used the data.”
These proofs of concept have helped to catalyze a sea change in the culture, and Malone says leadership is jumping at the opportunity to keep the momentum going by promoting wider implementation of JMP workflows. Already, he says, the company has documented a number of benefits related to JMP implementation, including:
- Data extracts are now standardized and accessible to everyone.
- Best practices for data analysis have been standardized across the department.
- Standardized display formats facilitate more efficient collaborations because graphics are familiar and therefore more quickly understood by everyone.
- Standardized best practices have led to better reporting and more effective meetings.
- Dynamic data exploration during meetings enables key stakeholders to ask questions and get answers in real time.
- More sophisticated data query and analysis has resulted in repeated time savings for every engineer, every day, across the Global Process Engineering Group.
- Teams no longer need to rely on a single source of data, thus improving business continuity.
- Increased analytics and visualization capability is now widespread.
From four days and an unknown to four hours and a decision
Automated workflows have already had a significant impact both on the pace of R&D innovation cycles and Johnson & Johnson Vision Care’s relationship with its customers. Prior to the rollout of standard JMP dashboards, process engineers had to go through an extensive data pull process that could take up to three or four days in order to investigate even a minor product quality issue. However, when Malone introduced a query and post-query script that gathered all the data necessary to disposition the material, join the data streams together, create graphics and produce an outcome – all in a single step – he says, “We went from needing four days to gather the data to actually having the data and a decision made in four hours.”
Malone recalls one instance in which, in anticipation of delays, the quality engineering group had come together to discuss the best way to communicate the timeline disruption to their customer while awaiting a product disposition. “We came into the meeting with a data-driven recommendation on how to disposition the product, and they didn’t believe us at first!” he recalls.
“They told us to go back and double check, but I said, ‘I’ll check it for you now.’ I just ran the script and query again in JMP… And they couldn’t believe it because typically [this process] would have taken us days,” he says – all thanks to a simple script.
This process has since become a standard internal monitoring service, with regular quality checks performed every week or two to see how the product and its parameters are behaving. “That has been a huge, huge benefit in terms of time,” Malone says. “Because obviously, if you’re holding a product while you’re waiting on a disposition decision, you can’t sell it. You’ve got to store it and you’ve got customers who are waiting.” These timelines are crucial in a marketplace increasingly moving toward just-in-time manufacturing. JMP, he says, has “proven its worth several times over.”
JMP is ‘future-proofed’
Change agents like Malone have invested heavily in supporting and encouraging colleagues to embrace data-driven approaches to engineering. And Johnson & Johnson Vision Care has invested financially in the evolution of its engineering organization. Malone therefore expects the returns on Johnson & Johnson Vision Care’s financial and time commitments to extend well into the future. That’s why it’s important to have invested in a tool that will grow with the organization.
“Everything we do to develop our capability with JMP seems to be future-proofed,” he says. “JMP keeps developing and increasing its capability… and its access to the data is never lost.”
Malone anticipates that as new lines are developed and new systems introduced, the need for broader data capabilities will only grow. And with its Early Adopter program and new version releases every 18 months, JMP is a partner that evolves alongside its customers, seeking active input from the engineers who use the software every day to solve real problems.
“Our time needs to be spent, as process engineers, making decisions,” Malone reflects. “JMP will give you back the time. It'll get you to the point where you're not just playing around with data but actually getting something out of it.”
