As manufacturers prepare for an explosion of data over the next few years, they will need to learn fast and embrace new analytical technologies, internal structures, and corporate cultures to turn that data into meaningful business-changing predictive insights, according to a new MLC survey. But many struggle with organizing around the data opportunity and developing ROI models. By Paul Tate
The amount of data generated by multiple M4.0 technologies is set to explode across the manufacturing sector over the next few years. Yet, many industrial enterprises are still learning how to deal with the data they already have. What’s more, a lack of effective tools, internal standards, control structures, and skills may hinder their efforts to deliver continual improvements in optimization, productivity, efficiency, and quality.
However, there are also clear signs that many manufacturers are already harnessing advanced M4.0 technologies and embracing new predictive analytical approaches to achieve the business-changing predictive results they seek. The overwhelming majority of manufacturing companies now believe that using and analyzing their data more effectively has become essential to future business competitiveness.
These are just some of the key findings from the Manufacturing Leadership Council’s first survey on M4.0 Data, conducted during March this year.
Data Explosion Ahead
Over the last two years a quarter of manufacturing company survey respondents say they have already seen their manufacturing data volumes double or almost triple in size. An additional 10% report they’ve experienced higher rises of between 200-500%, and another 12% have recorded even more dramatic increases above 500% (Chart 1).
And that’s only the beginning. Over the next two years, those manufacturers suggest that the rate of increase will rise significantly, with 38% of respondents now predicting increases between two and six times their current levels. What’s more, over a quarter of companies expect their data volumes to surge substantially by more than 500% over current levels in the next two years.
1 Massive Increase in Manufacturing Data Expected Over Next 2 Years
Q: What percentage increase in manufacturing data have you seen over the last two years ago, and what increase do you expect to see in two years’ time?
2 Over 50% Measure Data Value by Impact on Operational Performance
Q: How do you measure the value of the data in your organization?
3 Over 40% Have Corporate-wide Guidelines for Collecting & Managing Data Across the Enterprise
Q: Does your company have a corporate-wide plan, strategy, or formal guidelines for how data is collected and organized across the enterprise, including manufacturing operations?
4 70% Expect Over Half Their Data to Be Standardized in 2 Years.
Q: Considering the mix of new and legacy systems in most operations, what proportion of your manufacturing data is currently standardized around clearly defined corporate standards/formats, and what do you expect this to be in two years’ time?
5 But Data Governance Approaches Still Inconsistent: IT Heads & Joint IT/OT Teams Currently Dominate
Q: Who is responsible for data governance and strategy in your organization?
So far, the vast majority of those companies (54%) are measuring the value of those rapidly increasing data volumes in terms of the impact on operational performance (Chart 2), driving value through increases in productivity, efficiency, or quality. Only a handful (4%) has developed ways of measuring data value in monetary terms with an assigned dollar value.
Interestingly, a small number of companies (7%) are now beginning to measure data value against revenues from new data-driven services, a figure that may well increase in coming years as more manufacturing companies develop and deploy innovative data services around their increasingly intelligent products.
Perhaps the most unsettling result here, however, is that despite many years of investment, time, and effort in deploying digital technologies across their organizations, 30% of respondents admit that they have no measures in place so far to value the increasing volumes of data those technologies generate. Establishing a meaningful, data value-centric ROI formula for the digital era seems to be an issue that is currently being ignored, or at least overlooked, by many manufacturing companies. Or that they don’t yet have reliable models on which to evaluate or determine value.
What’s more, organizing and managing that data on a consistent and corporate-wide basis also seems to be proving tough to achieve. Over half the respondents say they do not yet have any corporate strategy, guidelines, or even a plan for the way that data is collected or organized across their companies (Chart 3), although 46% say they have already made efforts along this path.
Establishing consistent data approaches, of course, isn’t easy when many companies still have a disparate set of new technologies and legacy machines in use, often with different data formats and structures. But as data volumes grow, such corporate-wide initiatives are likely to become increasingly important as companies seek to control their data assets in a more consistent and unified way.
“The rapid shift towards more predictive insights will become increasingly important to achieving greater operational efficiency and resiliency.”
6 ERP, Quality, Shop Floor Systems Generate Most Data: Edge & Embedded Systems Begin to Emerge
Q: What are the primary sources of your manufacturing data today?
7 Productivity, Efficiency, and Quality Improve Most from Increased Access to Data
Q: How has the increase in manufacturing data helped you to improve your manufacturing organization?
8 Shift in Data Projects from Simple Analytics to Optimization & Prediction Over Next 2 Years
Q: What are your primary objectives for embarking on manufacturing data projects today, and what do you expect your primary objectives to be in two years’ time?
9 A Third Now Using AI to Analyze Manufacturing Data, But Spreadsheets & Shop Floor Systems Still Dominate
Q: What systems do you use to analyze the manufacturing data you collect?
10 Most Manufacturing Companies Are Still Learning How to Analyze the Data
Q: How would you rank your company’s ability to analyze the data from your manufacturing operations?
The good news is that over the next two years the industry may well see some rapid development in the way that companies seek to gain better control over data. The proportion of companies that expect to have over 50% of all their data standardized around clearly defined corporate standards or formats is set to double from around a third today, to 70% in two years’ time (Chart 4). The proportion of companies with only a quarter or less of their data in some kind of standard form today is set to dive accordingly, from 28% to just 4% in the next two years.
Steering those standardization initiatives will be the executives and teams that have been tasked with data oversight in the company. However, survey responses suggest that there is still a range of different approaches to who has governance and control over corporate data strategies in many manufacturing organizations (Chart 5).
Currently, most companies see the traditional IT function as the logical place for the coordination of corporate data, with governance responsibility placed under CIOs and IT VPs (23%), or perhaps joint IT/OT teams (18%). Some have even embraced the more recent concept of a corporate Chief Digital or Data Officer, although this title is still very much in the minority (7%). Again, there is unsettling evidence that many manufacturing companies have yet to tackle what perhaps in other areas of company assets would be considered a fundamental issue, with almost one in five companies (18%) admitting that no-one currently has overall data responsibility at all.
The vast majority of data now being handled by manufacturing organizations is generated by the company’s primary ERP systems (92%), quality control systems (80%) and shop floor systems (74%), although new technologies like edge computing systems (20%) and embedded systems in products (16%) are starting to make an impact (Chart 6).
In terms of operational impact (Chart 7), recent increases in manufacturing data have provided many companies with noticeable improvements in plant floor performance, particularly in productivity (77%), efficiency, (67%), quality (65%), and cost reduction (58%). However, it is also interesting that almost one in five companies say that increased access to data has also helped to spur innovation in their manufacturing organizations as well, adding another beneficial dimension to their M4.0 data journey.
11 Only Half Have a Process to Verify Data Accuracy and Quality Before Making Decisions
Q: Does your company have a process to verify the accuracy and/or quality of the raw data before decisions are made on it?
12 Over a Third Struggle to Capture the Right Manufacturing Data the Business Needs
Q: How would you rank your company’s ability to collect the right data the business needs from your manufacturing operations?
The Pursuit of Predictive Insights
Those desired performance improvements currently underpin many of the objectives that manufacturers say are motivating them to embark on new data projects in their organizations (Chart 8). Respondents reveal that today, the two primary objectives for starting new projects are to analyze shop floor data to understand their operations more effectively (70%), and then to use those insights to help optimize their operations for the future (56%).
But perhaps most significantly, they also suggest that over the next two years there will be a rapid shift to using more advanced analytical capabilities to be able to better predict key trends and potential events, up from 20% today, to a substantial two thirds of companies (66%) over the next 24 months.
That rapid shift towards more predictive insights will become increasingly important to achieving greater operational efficiency and resiliency but will also require companies to deploy ever more advanced analytical capabilities and technologies to be able to meet this goal.
Already, a third of companies say they are harnessing the power of artificial intelligence approaches to help them analyze the manufacturing data they collect, using both in-house AI resources (18%) and external AI partners (14%). However, there is still a lot of head room for change. The majority of companies still depend on more traditional shop floor analytics (60%) and even Microsoft Excel spreadsheets (68%) to analyze the manufacturing data at their disposal (Chart 9).
Most companies, however, are well aware that there is still a lot to learn along their data-driven journey.
Almost half of the survey respondents (47%) admit that their companies are still learning how to analyze data from their manufacturing operations effectively (Chart 10). Less than one in five (18%) believe their companies are currently “very capable” of analyzing the data they have, and 27% suggest they are only “somewhat capable” at this stage.
In addition, only half of the survey respondents say they already have processes in place to verify the accuracy and quality of their data before they start to make decisions based on it (Chart 11), and over a third of organizations say their companies still have only a “low” ability to actually collect the right data the business needs from its manufacturing operations in the first place (Chart 12).
13 Data-driven Operations Increase Collaborative Decision-Making and Use of Data Performance Metrics
Q: In what way has your organizational structure adapted to optimize increasingly data-driven operations?
14 Lack of Common Formats, Easy Access, Data Capture Systems, and Analytical Skills Are Main Barriers to More Data-driven Decisions
Q: What are the most important challenges or obstacles hindering your organization from making more data driven decisions?
15 Manufacturing Data Essential to Future Competitiveness
Q: Looking forward, how important do you think manufacturing data will become to your competitiveness as a future business?
Alongside these areas of concern, there are also a number of other obstacles respondents highlight as potential hinderances to making more data-driven decisions (Chart 14). Most significant is the heritage of legacy systems in many companies which has created a mish mash of different systems across their production networks, producing data in different formats and making it more difficult to consolidate and unify their data strategies (46%). Respondents also point to difficulties in easily accessing the data they need (40%), a lack of the right systems to capture all the data they would like to have (38%), and a lack of skills to be able to analyze their data effectively (33%), as their top challenges.
Changing the Organization
Nevertheless, an increasing focus on developing more effective data-driven operations is already having an impact on organizational structures and incentives systems in many manufacturing organizations.
For example, around 45% of companies say their approach to decision making has now become more collaborative as their organizations have adapted in their efforts to better optimize increasingly data-driven operations (Chart 13). Fifty three percent of companies have also established dedicated data teams at either a functional or corporate level. And around 35% of companies have specifically embedded data targets into their performance metrics for their employee teams.
“An increasing focus on developing more effective data-driven operations is already having an impact on organizational structures.”
What is clear, though, is that the vast majority of manufacturing companies now realize that overcoming these potential obstacles and creating better ways of collecting, managing, analyzing, and exploiting manufacturing data is going to make a big difference to the future success of the organization. An overwhelming 73% of respondents now accept that manufacturing data will become “essential’’ to future competitiveness (Chart 15), and another 27% believe it will at least be “supportive”.
Not one respondent claimed manufacturing data will have “no impact” at all.
That universal awareness of the growing importance of manufacturing data to future business competitiveness is perhaps the most significant result of all. However well prepared many companies may be today, there is now a powerful driving force across the manufacturing sector to pursue more predictive insights into operations and products to help make faster and better data-driven decisions. In many ways that constitutes a declaration of industrial intent to determine a better data destiny for manufacturing in the years ahead and that data mastery will be an essential core competency for the future. M