At-06079 Promv WP Updated 060420
At-06079 Promv WP Updated 060420
At-06079 Promv WP Updated 060420
White Paper
Regardless of the catalyst, the need for rise in food allergies, changes to regulatory
continual change is the ironic constant for most requirements, and an increased focus on food
global industries. This is certainly true for the safety, food manufacturers have a complex
food and beverage processing and engineering set of challenges to navigate when it comes to
industry, which as of late 2019, was estimated what they produce and how they compete for
to grow to $4.1 trillion by 2024.1 With the the attention and wallets of customers.
world‘s population expected to balloon to
9.3 billion by 20502 and agricultural production Digitalization in Food and
needing to grow by at least 70%3 in order
to feed all these people, it’s easy to see why
Beverage Processing
companies throughout the industry might be Digitalization is a buzz word in many
feeling immense pressure. As the population industries. With the success of food and
surge directly impacts the amount of food beverage companies hinging on customer
needed and consumed, shifting economic tastes, evolving preferences and speed to
conditions and the need for manufacturers to market, the ability to leverage digital
keep up with changing consumer demands are technologies and create efficiencies enables
driving big changes in the types of products new competitive advantages for businesses
being produced. implementing innovative strategies, especially
those related to the “autonomous plant.”
What’s being manufactured is also partially Specific gains in the areas of improved supply
impacted by an increased desire to eat ‘cleaner chain processes, streamlined production and
and healthier’ items. For example, plant- simplification of compliance tracking are the
based protein sales increased by 24% in 2017, most frequent examples cited on the benefits
which most likely shifted the focus of some of an accelerated digital strategy.
developers.4 But this must also be balanced
against other data, such as one study that For many companies, it can be complicated,
found that Asian protein consumption will even overwhelming to know which steps to
grow by 128% by 2050.5 Considering the take first, the technology to have in place
2
and how digitalization will actually impact Food Safety Modernization Act (FSMA) and Without understanding the exact variances
the bottom line. Some businesses may be a new trend toward ingredient traceability as a (not always visible to humans), production
uncertain which products could best align key component to building consumer trust. To cannot easily forecast the outcome or effectively
with a comprehensive digitalization strategy. better understand digitalization’s role, we also compensate, potentially resulting in an
Automation of data collection for compliance, need to examine the factors impacting quality. expensive defective batch.
lean manufacturing and asset optimization
are common efficiency examples. There are Hidden Factors Impacting Let Your Existing Data
also successful instances of digitalization that
enable new services to improve the customer
Quality Improve Your Confidence
experience on the delivery side. But what role Even with specifications defined within a in Quality
can digitalization play in overall product product formulation, minor variations in
quality as a competitive advantage for the processing may still arise. These differences Particularly in the food and beverage industry,
food and beverage industry? can steer production unknowingly off course. mountains of data exist to comply with
Despite traditional means of recording process food safety tracking requirements. However,
Food processing on a large scale is particularly data, quality compromises might arise from: information such as maintenance work orders
complex. While quality control has defined and compliance reports typically resides
processing parameters of variables, including elsewhere in an organization, potentially
• discrepancies in raw material properties
time, temperature and good manufacturing separated from related material.
• variations in procedures
practices (GMPs), there are stricter regulatory
requirements resulting from sources like the • changing environmental conditions
3
This division of data means that past efforts to mine it may not have outcome. With many data visualization and analysis techniques, it may
yielded notable results. As digital transformation efforts increase, some still be a daunting challenge for operators, engineers and analysts to sift
companies are looking to this historical information to tell a story about through all the data patterns and identify how best to adjust processes.
how they can improve quality in the future.
What is the first step in utilizing past production data to provide insight? Extracting Meaningful Data Increases
With traditional data analytic methods, many businesses would be
Quality of Production Supply
advised to hire statisticians and data scientists to perform complex
analyses. However, in addition to being unrealistic for most organizations, The more quickly you can turn raw data into actionable insights, the better
this approach neglects the other subsets of data in companies’ hands: the – especially if you can do it without teams of new resources. Multivariate
“hidden factories” or inherent knowledge of their site operators and others analytics software can help solve your process and product quality issues,
who are intimately familiar with the ins and outs of the process itself. which could likely result in greater customer satisfaction. These tools can
While they may not have been the creator of the product itself, they are help optimize production by:
intimately familiar with how to execute on the product formulation given
the surrounding circumstances. • reducing off-spec product
4
Aspen Technology offers a comprehensive suite The customer’s first attempt at reducing
of tools to effectively collect, organize and reliance on lab testing involved instituting
analyze batch and continuous process data, an intermediate, mid-process test. This test
empowering a step change in your required less time to complete and was
organization’s ability to take hold of the done more often, but results did not correlate
promises of Industry 4.0. to final product quality. Because the test did
not ultimately act as a good predictor to
Three Use Cases: yield 100% product quality, this process
was stopped.
Taking on Big Data
Challenges with Confidence Because of a lack of success utilizing traditional
with Aspen ProMVTM methods, a digitalization solution was adopted
applying Aspen ProMV to the process data
Here are examples of food and beverage from the plant historian over a four-week
companies, and organizations experiencing period. The customer was able to assess
similar challenges, taking advantage of Aspen roughly 150 different variables to model ‘good’
ProMV to improve quality in real time and process conditions. With this model complete,
optimize production. Aspen ProMV Online was utilized to
continually monitor behaviors in process
1. Predicting Quality in Real Time
variables that indicated the process might
The process engineers at a global food
produce off-spec product process variability
processing and commodities trading
was detected, a real-time alert was received.
corporation had an issue on their hands. While
Next, Aspen ProMV was applied to assess the
quality challenges were not always persistent,
alert and prescribe actionable insights on how
there often was no clear solution to root cause
best to correct the issue.
them when they did occur. Lab tests utilized for
final product quality often lagged production by
two to three days, making it difficult to trace or
correct the causes of bad quality in real time.
5
Adopting Aspen ProMV as a predictor of The customer decided to adopt Aspen ProMV ever noted or assessed. Installing a dedicated
quality-armed operators, plant management to help determine the root cause of the quality cooling water line on production line #2
now had the ability to rapidly determine the variance. Approximately eight months of data resolved the issue and eliminated the root
location of a root cause by predicting lab and 1,100 data points were analyzed in two cause contributing to the quality variance,
testing results in real time. As a result, the days. These data points included information saving the site over $400k in operating costs
customer saved more than $250k per year about the entire facility, and were not limited to per year.
(USD) in off-spec product. the production line that was experiencing
the issue. 3. Maintaining Quality While Increasing Supply
2. What If the Needle Is in a Different Haystack? Chain Flexibility
An international agricultural products company Over a three-day period, results were reviewed An international food manufacturer traditionally
with three production lines running in parallel and the Aspen ProMV model fine-tuned to tested incoming raw materials as its primary
was experiencing random variances in quality determine the quality variance root cause. method for predicting overall product quality.
on a dedicated production line. Operators Once all contributing factors were assessed in This methodology was resulting in unreliable
noticed a key quality parameter randomly Aspen ProMV, results showed process outcomes and an unacceptable level of
changing over time — contributing to final conditions on an entirely different production line off-spec product.
product quality on production line #3. As a (production line #2) were the biggest
result, the company had over $400k per year contributors to the quality variance. Although it Having multiple raw material suppliers for each
in unusable product over the course of several was known that the two production lines ingredient of a specific product, the customer
years. All investigation efforts attempted over shared a cooling water source, no linkage was faced with a dilemma on how to best pre-
the years were unsuccessful. between quality output of the two lines was determine the outcome of final product quality.
6
With the availability of historical data of the product’s raw material lots and variable processing
conditions, Aspen ProMV was utilized to develop a comprehensive data model that correlated
these two factors. The data associated with the model revealed two raw materials had no
significant impact on final quality while three others did. Therefore, the manufacturer was able to
determine where to focus their raw material specification efforts to eliminate future off-spec
product and adjust their manufacturing process to improve overall quality.
By using Aspen ProMV, the customer was able to scale the same modeling technique across their
entire product line. This helped them avoid raw material combinations that would have led to
future poor final quality — ultimately contributing to increased product shelf stability while
avoiding product recall potential.
“Global Food Processing Market Opportunity, Trends, Forecast and Competitive Analysis Report 2019 with Profiles of Nestle,
$
1
PepsiCo, Archer Daniels Midland, Unilever, Anheuser Busch Cision PR Newswire, 7 October 2019
2
https://www.rentokil.com/blog/food-processing-trends/#.XgYoz-hKg2w
3
“7 Global Trends Impacting Food Processing.” Jack Lyons, Rentokil, 16 August 2019.
4
“3 Trends in the Food Processing Industry.” Acuity Insurance, 31 July 2019
5
https://www.rentokil.com/blog/food-processing-trends/#.XgYoz-hKg2w
7
About Aspen Technology
Aspen Technology (AspenTech) is a leading software supplier for optimizing asset performance.
Our products thrive in complex, industrial environments where it is critical to optimize the asset
design, operation and maintenance lifecycle. AspenTech uniquely combines decades of process
modeling expertise with machine learning. Our purpose-built software platform automates
knowledge work and builds sustainable competitive advantage by delivering high returns over the
entire asset lifecycle. As a result, companies in capital-intensive industries can maximize uptime
and push the limits of performance, running their assets safer, greener, longer and faster. Visit
AspenTech.com to find out more.
www.aspentech.com
© 2020 Aspen Technology, Inc. AspenTech®, Aspen®, aspenONE®, the Aspen leaf logo, the aspenONE logo
and OPTIMIZE are trademarks of Aspen Technology, Inc. All rights reserved. AT-06079