adaptive-digital-factory-whitepaper

THE ADAPTIVE
DIGITAL FACTORYTM
The Internet of Everything in End-To-End
Manufacturing and Product Lifecycle Management
Carolyn Rostetter
Dr. Setrag Khoshafian
Kenty Adams
Foreword By: Bruce Williams

WWW.PEGA.COM1
FOREWORD 2
.01 WHY THE ADAPTIVE DIGITAL FACTORY 3
.02 SMAC’T DIGITIZATION TRENDS 4
.03 INTERNET OF EVERYTHING FOR MANUFACTURING 5
.04 DYNAMIC CASE MANAGEMENT 6
.05 IT AND OPERATIONAL TECHNOLOGY INTEGRATION 7
.06 COMPELLING USE CASES 8
.07 THE OODA LOOP IN MANUFACTURING 9
.08 BIG DATA BECOMES THING DATA 10
.09 DIGITAL TRANSFORMATION FRAMEWORK 11
.10 DIGITIZING THE SUPPLY CHAIN 12
.11 INNOVATIONS ON THE SHOP FLOOR 13
.12 THE SHOP FLOOR 14
.13 THE END OF MONOLITHIC MANUFACTURING? 15
.14 DYNAMIC CASE ORCHESTRATION 16
.15 SECURITY CONCERNS 17
.16 MASTER DATA AND MASTER POLICY MANAGEMENT 18
.17 EVOLVED PRODUCT LIFECYCLE MANAGEMENT (PLM) 19
.18 DEVICE-DIRECTED WARRANTYTM
20
.19 DIGITAL PRESCRIPTIVE MAINTENANCE 21
.20 ALIGNING CUSTOMERS AND MANUFACTURERS 22
.21 INDUSTRY 4.0 AND THE INDUSTRIAL INTERNET 23
.22 THE DIGITAL TRANSFORMATION OF MANUFACTURING 25
REFERENCES 26
CONTENTS

WWW.PEGA.COM2
Digital Transformation has arrived in global manufacturing, and “manufacturing” as we’ve known it – in the value
streams of making things as well as the industrial systems that surrounds them – will never be the same.
In The Adaptive Digital Factory, my colleagues at Pegasystems and Jabil Circuit have outlined in detail the impacts
of the digitization phenomenon now flooding across global manufacturing. This treatise addresses the full range
of emerging technologies and how they are transforming the very nature and performance characteristics of the
manufacturing industry. Digitizing the value stream and aligning the digital world with physical devices enables new
ways of working across the product life cycle – from design to sourcing, through the supply chain, across the shop
floor and into the all-important aftermarket. The technologies themselves – social, mobile, big data, analytics, cloud
and of course the Internet of Things – are re-defining everyone’s roles and operating models. It’s clear that the
fundamental landscape of the industry is in transformation, and that within the next decade, many manufacturers
will find themselves managing operations so new and different as to be virtually unrecognizable from the past
century of industrial manufacturing.
At Pega, we are very excited by how well we are aligned and suited to support the convergence of these industry and
technical forces. Enabling this transformation requires enterprise software built with unprecedented levels of ease,
utility and flexibility. Applications must be business-centric, process-oriented and robustly adaptive to continuous
change. They must now include all forms of actors, including people, systems and things. And they must support
both established and emerging practices such as Dynamic Case Management, Lean Six Sigma, and Industrie 4.0.
Pega technology and applications natively support all of this at the speed and scale required.
The Adaptive Digital Factory is a comprehensive survey of the drivers, the methods, technologies and innovations at
play across the manufacturing landscape. The authors also include example use cases and specific applications. It’s
presented in an easy-to-read, one-page-per-topic format that serves both as a primer and a reference.
Manufacturing – of devices, materials and energy – represents over a third of the world’s total economic output.
The changes being wrought throughout these industries from digitization represent one of the great macro-scale
movements of our time. The authors of The Adaptive Digital Factory are experts in this transformation. They have
smartly compiled an overview for quick consumption and provocative consideration for the profound implications.
BRUCE WILLIAMS
VP, Manufacturing
Pegasystems
Foreword

WWW.PEGA.COM3
The Internet of Things,[1]
also referred to as the Internet of Everything, the Industrial IoT,[2]
Machine to Machine, or
“connected devices,” has the potential to transform and disrupt traditional manufacturing models. The connection of
devices through IoT provides transformational opportunities for managing the shop floor, product lifecycle, supply
chain, and the manufactured devices themselves.
DEFINING THE DIGITAL FACTORY
The revolutionary digitization of decisioning for end-to-end business processes and dynamic cases[3]
is an
essential enabler of digital factories. By leveraging the capabilities of the Internet of Things, manufacturers remain
connected to edge devices 24/7 and, through that connection, their end consumers.
Consider the following interrelated and interdependent dimensions for digital factories.
OPTIMIZED FACTORY ASSETS
In traditional manufacturing, the majority of assets on the shop floor are not interconnected. As a result,
the function and status of those assets are, for all practical purposes, invisible to the manufacturer. With IoT
connectivity, all assets and machines can be connected, monitored, controlled, and optimized.
PRESCRIPTIVE MAINTENANCE[4]
The digital factory’s connected machines, whether on the shop floor or manufactured devices, can be maintained
proactively and prescriptively. This prescriptive approach is a key advantage of the IoT model in terms of
maintenance, monitoring, controlling, and updating of connected products.
LIFETIME VALUE OF MACHINES
IoT leverages Big Data[5]
analytics to optimize the lifetime value of manufacturing machines. Furthermore, with
digitized and connected manufacturing, the lifetime value of manufactured devices also improves.
BIG DATA PREDICTIVE MODELING
Manufactured devices continually generate enormous amounts of sensor data. The digital factory mines this data for
potential patterns, and identifies the preventive maintenance actions required to avoid potential incidents or failures.
3D PRINTING[6]
AND IOT
The advent of affordable 3D printing will allow organizations to manufacture connected products cost effectively
and engage multiple innovators through a connected fabric of manufacturers.
Digitization enables the digital factory
manufacturer to innovate on the shop
floor in real-time, as well as throughout
the lifetime of connected products.
With digitized dynamic cases, Things,
people, analytics, and process
connectivity streamlines manufacturing
from suppliers to consumers, and
allows for significantly increased
adaptability.
.01 Why the Adaptive Digital Factory?

WWW.PEGA.COM4
Considering Social Networking, Mobile, Analytics, and Cloud Computing
Digital factories must take advantage of other digitization trends including social networking, mobile, analytics, and cloud
computing. The phrase “SMAC’T”[7]
references these important trends as well as the IoT (which supplies the “T” for Things[8]
).
IMPACT ON MANUFACTURING[9]

Let’s examine how these four digitization trends provide value to manufacturers.
SOCIAL NETWORKING
Social networking has a positive and sometimes challenging impact upon manufacturers. The customer’s[10]
social
postings may support or damage manufacturers. On the other hand, social collaboration aligns customers with
dealers and distributors, as well as manufacturing service and production teams.
MOBILE
Mobile omni-channel and omni-device interaction allows manufacturers to seamlessly initiate and complete
automated casework end-to-end through mobile devices. Omni-channel refers to the consistency of the customer
experience across different channels. Omni-device refers to consistent connectivity, maintenance, and optimized
experience with the connected device.
ANALYTICS
Analytics provides manufacturers with necessary tools to mine the hidden gems buried in overwhelming amounts
of data. Once identified, manufacturers can act upon key information pertaining to their products, services, and the
processing of manufacturing materials. Emerging trends in Big Data and real-time analytics enable manufacturers to
optimize their manufacturing, maintenance, and other processes.
CLOUD COMPUTING
Cloud computing allows the entire
manufacturing value chain (from
consumers to manufacturing
management to employees) to
access any phase of the product
lifecycle. Cloud computing services
delivered over the Internet
through networks, servers, storage,
and business applications offer
convenience and on-demand access.
When the power of IoT connected
devices is added to the capabilities
of mobile, social, cloud, and analytics,
traditional manufacturing can
embrace the transformation to
Adaptive Digital Factories.
.02 SMAC’T Digitization Trends

WWW.PEGA.COM5
The Most Significant Digitization Trend
The Internet revolution, which connected people through digital devices such as wearables, unmanned cars,
and smart homes, is poised for further expansion through a massive network of sensors and actuators. For
manufacturing, the revolution continues with the advent of connected assets and manufactured devices that are
continuously connected, monitored, and optimized by the adaptive digital factory.
Manufacturers have begun to use connected devices on the shop floor, and now must look to the IoT[10]
to
access uniquely addressable physical devices over the Internet. They need to work with and create Things that
increase the connectivity between the manufacturer and consumer. These connected devices can have on-board
diagnostics software that manufacturers can use for monitoring, control, and management.
WHAT ARE THINGS?
Things by themselves have little value. However, when considering the health and maintenance of Things, it is
important to take a holistic view of how these connected devices fit into the IoT. For that view, Cisco describes IoT
as Internet of Everything[11]
with a holistic view:
1. Things: Connected devices, each with a unique addressable URL over the Internet.
2. People: Participants in end-to-end processes or dynamic cases, in partnership with Things.
3. Process: At the most basic level, coordination of multiple activities or tasks for a business objective. The
participants in processes will include people, Things, enterprise applications, and business partners.
4. Data: Generated by Things at an even greater rate than people or applications, IoT data needs to be
analyzed and responded to, often in real-time.
There are core relationships between and among these four entity types. The process defined here is a generic
term referring to any type of process, whether structured or dynamic (including dynamic cases).
.03 Internet of Everything for Manufacturing

WWW.PEGA.COM6
The Optimized, Automated, and Adaptive Digital Factory
Shop floors involve complex processes, each of which may encompass multiple milestones, tasks, and business units
in the workflow required to handle and produce a manufactured good. Manufacturing aftermarkets (e.g., maintenance
and support) also depend upon and involve complex processes[12]
that reach beyond the confines of the factory and
present challenges for coordinating and managing these interconnected yet distinct resources.
Dynamic case management’s sophisticated end-to-end support of connected devices proves it’s the optimal
methodology for facing the challenges (and opportunities) of the digital factory.
DYNAMIC CASE MANAGEMENT CAPABILITIES
DCM capabilities address each of the complex components inherent in IoT cases, including:
ƒƒ Case Stages and Objectives: Create, manage, and report on the processes and Cases of Everything. The
model-driven design of cases is achieved by identifying the milestones (stages) of each case. Pega’s manufacturing
solutions easily map value streams spanning customers, manufacturers, suppliers, and field through stages in DCM.
ƒƒ Things as Participants [13]
in DCM. Case stages and processes are acted upon by numerous participants,
including manufacturing, back-office staff, field service providers, customers, and Things. DCM allows multiple
entities (humans or Things) to view, review, work, and subscribe to a case and its processes.
ƒƒ Monitor and Improve: DCM portals and real-time reports allow business activity monitoring of the
manufacturing case stages. Users track the progress of work throughout each stage of the case. Pega Business
Intelligence Exchange allows data to be exported to a data warehouse, providing complete DCM visibility
throughout the enterprise.
ƒƒ Case Collaboration—Pulse: Collaboration takes the form of instant messages, files, and URLs that can be
shared with other users within a work group. Posts are designated as public or private.
ƒƒ Case Data (Properties): Cases and processes may employ different data models from a variety of sources,
including user input, external systems, the result of business rules, and calculations. This data may be
propagated throughout all of the processes in the case, which enhances case and process functionality.
ƒƒ Dynamic and Ad-Hoc Tasks: Tasks are assigned to humans, Things, or automated agents (bots, robots, etc.) based
on identity, role, skills, or a combination of all three. Workers and managers can dynamically introduce ad-hoc tasks
and even discover the need for new processes in the context of an individual case. Dynamic and ad-hoc task capability
supports inevitable variations in work, smarter work processing, and social collaboration to achieve business objectives.
Dynamic case management supports end-to-end connectivity from the edge devices—on the shop floor, in the
field, or with the consumer—to the rest of the enterprise.
.04 Dynamic Case Management
Sub-Cases Case
Content
Tasks Case
Events
Processes
Case DataCollaboration Case
Stages
1 2 3
Case
Workers
Case
Business
Objectives
Decisions
Dynamic Case Management

WWW.PEGA.COM7
Integrating Operational Technology Systems with IT[14]
Historically, manufacturing has had two different domains: operational and informational. Operational
Technology (OT) concentrates on assets or devices that need to be sensed, monitored, and controlled in real
time, whereas IT focuses on managing the portfolio of applications across the enterprise.
Today, the physical world and the Internet are converging at a rapid pace because businesses want to integrate
data from MES, SCADA, PLM, and process control systems for new types of analytics and intelligence.
OT MEETS IT IN THE ADAPTIVE DIGITAL FACTORY®
In the Adaptive Digital Factory®
devices are becoming increasingly smart and automated for monitoring,
maintenance, and continuous improvement. These devices, which may be human- or computer-controlled, are
connected to the full suite of production IT applications. In order to create connections among the organizational
silos and improve the flow of value to customers, process flows must be digitized and operationalized.
There are emerging reference architectures and communications standards for OT-IT integration. One of these is
the ISA-95[15]
standard, which is designed to integrate IT enterprise software with manufacturing shop floors. The
following illustrates the levels and the various systems involved.
.05 IT and Operational Technology Integration
Sourced from: IoT World Forum Reference Model
http://www.1otwf.com/resources

WWW.PEGA.COM8
Connected Devices for Manufacturing
As stated earlier, Things by themselves have limited value. There are three key Process of Everything (PoE)[16]
transformational use cases for manufacturing: a focus on intelligent digitized processes, sensing and responding
to exception events, and operationalizing big data generated by Things.
DIGITIZING PROCESSES
Elevating the focus from technology stack to intelligent digitized processes
A key requirement for truly successful digital transformation of IoT is the end-to-end digitization of processes. A
process has inputs, orchestration of tasks, and business outcomes upon its completion as output. Traditionally,
process automation orchestrated humans, business partners, or enterprise applications. That landscape of process
participants is changing with the rise of Things.
Consider the coordination of the supply chain involving the supplier, OEM manufacturer that assembles the connected
Things, distributors and potentially retailer, or dealer. The participants include staff from different organizations on the
supply chain as well. Things can be queried for sensor data, or they can be ordered to carry out tasks. Autonomous or
semi-autonomous Things are becoming active participants in business processes.
SENSING EVENTS AND DIGITIZING CHANGE
Handling events and digitizing change
One of the most pervasive use cases for Things is sensing (through IoT sensors) an exception event and then
activating a digitized end-to-end process to respond and resolve it. This happens when there is a mechanical or
software failure on a device. The Thing autonomously senses and then either directly, or through a brokering layer,
activates an exception process. This typically includes monitoring back-office and field workers to respond and
resolve it. There are many policies in play when executing end-to-end digitized processes with IoT.
TRANSFORMING THING DATA INTO DIGITAL PROCESSES
Operationalizing Thing data analytics
Often, it is not merely an individual event that instantiates a process. Big Data will eventually become “Thing Data.”[17]
Over
time, Thing data (and digital processes) will be stored and analyzed. Then, through analytic techniques, predictive models
can be mined for information that will become a leading source of Big Data.
Digital Transformation discovers patterns in Thing Data and then operationalizes those patterns in digital processes.
Sometimes as these events occur, they must be correlated in real-time analytics. Based on either predictive analytics
models or real-time event correlation, a digital transformation solution can take action or suggest the next-best-action
to humans or smart Things (e.g. robots).
.06 Compelling Use Cases
Operationalize,
Process and
Respond
OutcomesPattern
Detection with
Real Time
Analytics
EVENTS

WWW.PEGA.COM9
Strategic Decisioning in the Digital Factory
The OODA[18]
Loop stands for observe, orient, decide, and act. It’s also
known as the Boyd Cycle, named after retired USAF Colonel James Boyd
and his approach to gaining strategic advantage over enemies in battle.
OODA also correlates to the Lean manufacturing P-D-C-A Cycle
(Plan–Do–Check–Act). Both OODA and PDCA are concerned with how
decisions are made. Simply put, they stress the need to observe data
and events, determine the meaning, filter out the noise, reorient as new
information comes in, and take action all in a continuous iterative loop.
THE OODA MODEL
When employed in the digital factory, the flow of actions and reactions in the OODA model is a process that includes
humans, machines, and data from things. In a tough market challenged by low operating margins, the ability to
think and act more quickly than the competition creates a distinct advantage. In fact, improved supply chain
management and lower operational cost are the key drivers of many current PoE initiatives in the manufacturing
and high tech industries.
Let’s examine how the components of OODA apply within the digital factory.
OBSERVATION: Observation corresponds to either the experience or intuition of a digital factory knowledge worker
or the collection of the data (increasingly Thing Data) for discovery. Sources of this data include Things, processes, and
enterprise applications.
ORIENTATION: Orientation emanates from the knowledge and insight within a particular context. The orientation
comes from capturing and digitizing the expert’s knowledge or through business rules. Alternatively, orientation can be
mined through discovering models (e.g., predictive analytics) from data.
DECISION: From the observations and orientation, a prioritized set of decisioning options can be presented. The user
or the system needs to pick the course of action, which in most cases will be the first option or best action from a list of
potential actions.
ACTIONS: These are actions taken in the implementation and application of a decision in a particular context within
the digital factory. Here the user acts upon a prioritized list of decisions, also known as the Next-Best-Action, in a
particular context or situation.[19]
TYPICAL USE CASES
In fact, the majority of IoT use cases that our customers focus on are:
ƒƒ Detecting what is happening whether from streaming data from machines and devices or from the
occurrence of complex or anomalous events.
ƒƒ Automating the triage, diagnosis, and recommended actions with the highest propensity to fix the problem
(or better yet to prevent the problem from happening in the first place).
ƒƒ Executing on actions to resolve the problem, typically using Dynamic Case Management. Over time as the
system detects and resolves enough problems, it eventually becomes self-learning and automatically begins
the next iteration of the continuous OODA Loop.
.07 The OODA Loop in Manufacturing
OODA
Loop
Act
Observe
Orient
Decide

WWW.PEGA.COM10
Building a Data Infrastructure for Adaptive Digital Factories
With all the data that is accumulating from all the devices, networks, people, links, machines and so on, there is no wonder
we term this accumulation Big Data.[5]
In just a few years, we will have access to petabytes of structured, unstructured, and
semi-structured data. With that quantity of data at our fingertips, it would be unwise to define Big Data as a repository.
It would be the best way of losing sight of the hidden insights and knowledge that the data is providing.
BUILDING A THING DATA/BIG DATA INFRASTRUCTURE
One of the most effective approaches to managing Big Data is building the repository use case by use case. By
taking a use case approach, it is possible to understand and know our data, its nomenclature, its function, and its
transformational needs. Additionally, analyzed use cases that produce positive outcomes (e.g., concrete action) trigger
new questions and birth new use-cases.
Combined, the accumulation of use cases, related insights, and the resulting actions all contribute to building an
effective Big Data platform. This Thing Data/Big Data platform develops from an intelligent, ordered, and structured
methodology, which allows decentralized analysis teams to leverage the previously worked use cases.
One additional significant technical advantage of employing a use case methodology—it facilitates the scaling of the
infrastructure. There is no need to commit significant resources to developing an immense infrastructure at the onset.
Rather, the platform starts small and then expands as the need increases.
In the manufacturing world, data is generated by each machine, by each interaction between machines, by each
sensor within/amongst machines, testing machines, controlling systems, operators, engineers, and so on. Today,
only a fraction of that data can analyzed.
As products become more and more intelligent, they produce valuable data. Now that systems, tools, and
repositories are friendly enough to consume all this data, the manufacturing world must acquire the tools to access
this untapped knowledge.
Consider the near future: a simple production line will soon be able to communicate, send data, and save it in the
cloud in order to build use cases from it. One possible use case could be decreasing the Takt time of the line, an action
that has clear benefits in manufacturing. In this example, data about the production times might be captured every
second and sent to a SaaS (Software as a Service) repository like a Hadoop[20]
environment. The data would then be
analyzed using sophisticated statistical tools, and a concrete action would be taken to change pricing requirements
based on knowledge gained from the analysis. In turn, this use case might provide new insights on the same problem,
bringing new data to light that informs new analysis, producing new actions, and delivering new business value.
.08 Big Data Becomes Thing Data
Action
Business
Value
Business
Purpose
Insights/
Analytics
Use Case by
Capability/Service
BIG DATA

WWW.PEGA.COM11
Strategic Harnessing of Valuable Data
The manufacturing industry has come a long way from the mechanical processes introduced by the Industrial
Revolution,[21]
and has enjoyed the efficiencies of decades of computerized manufacturing processes. Even greater
benefit is promised, however, as we move into the era of digital information revolution. Valuable data is proliferating, and
manufacturing must harness it to reach the next level of competitiveness.
It is clear that the manufacturing industry will transform as the result of the inevitable digitization of many production
processes. The following describes one way of facilitating this transformation.
ƒƒ Identify: First, we must identify the use case and decide, for example, whether we want to optimize a
particular process, manage another, or monetize a third.
ƒƒ Enable: Something must be designed to produce data so that information and insights can be produced.
This data may be inherent to the systems being used, or may need to be implemented in order to produce
that data (install, configure, connect, and disseminate). Examples of this “enabling” are different types of
sensors including RFID, WIFI, mobility devices, robots, and so on.
ƒƒ Capture: Devices produce data, but without interpretation, it is unintelligible. To capture data, applications
are developed on top of these devices to produce outcomes. At this level, applications capture the raw data
emanating from devices.
ƒƒ Transform: To analyze the captured data, it often needs to be transformed so that it is either coherent with
other sources of data, joined to other data, or loaded to a specific data repository. This stage performs the
extraction, transformation, and loading (ETL) required so that the data can be consumed for analysis.
ƒƒ Store: As data increases exponentially, it needs to be stored to be consumed by use cases. This repository
aspect of the transformation takes into consideration the type of data (structured, unstructured, semi-structured),
volumes, ease of access, security, recovery, and others.
ƒƒ Analyze: The analysis phase takes into account internal and external macro/micro economic considerations
and the acquired knowledge of the data, and then applies statistical or mathematical modeling to gain
insights in order to take action.
ƒƒ Action: Having gained informed insights, action is taken and new use cases are defined.
.09 Digital Transformation Framework
Identify
Enable
Capture
TransformStore
Analyze
Action
(Manage, monetize, extend & operate opportunities)
(Data from sensors, RFID, WIFI, mobility, & robotics)
(Applications and Systems)
(API’s , Data services (ESB)
& Integrators)
(Hadoop Distributed File System (HDFS))
(Statistical modeling,
Machine learning &
Data Science)
(BPM, DCM, Decisioning,
Event management,
Complex Event processing, APIs)
Digital
Transformation
COE

WWW.PEGA.COM12
Improving End-to End Performance with Dynamic Data Mining
One needs look no further than the evolution of supply chain capabilities to see how manufacturing is becoming
digitally connected. Digitization of the supply chain through monitoring, connectivity, and end-to-end automated
dynamic cases has provided many new opportunities to optimize the supply chain[22]
processes.
With the emergence of additive 3D printing[23]
for “social” manufacturing, optimization within supply chain management
is critical. Companies rely on the timely supply of material or 3D components in order to produce at increasingly
demanding service levels. Any disruption to the supply chain has a negative impact.
From a financial perspective, costly disruptions may result in overstocking, underproduction, waste, and, ultimately, the
visibility into the supply chain from end-to-end. For that reason, disruptions in the supply chain represent a significant
risk for the company.
In order to minimize that risk, companies must identify the sources of the disruptions.
.10
DISRUPTIONS IN SUPPLY CHAIN
Some of the critical sources of supply chain disruptions
include:
ƒƒ Weather (e.g., hurricanes)
ƒƒ Civil unrest (e.g., protests)
ƒƒ Labor union challenges (e.g., ports labor dispute)
ƒƒ Fuel prices (e.g., unwarranted hikes in the rates)
ƒƒ Vendor market status (e.g., hostile takeovers of
suppliers)
In order to minimize disruption risks, data is mined and
transformed into information used to take mitigating
actions. These mined actions are automated and executed
through end-to-end dynamic cases spanning the entire
supply chain.
MITIGATION BY PREDICTION
Some of the sources of data for analytics, decisioning, and especially operationalizing in the supply chain
processes include:
ƒƒ Social media: Useful for trending on possible civil unrest or labor union issues
ƒƒ Text analysis: Used to analyze social media input, define trends in communications, preferences, and ideas
(e.g., millennials versus baby boomers)
ƒƒ Risk prediction: Derived from information about weather, climate, or geological predictions
ƒƒ Machine learning: Algorithms that learn, adapt, and become more accurate at predicting with more information
The data from these sources are also supported by sensor data from the transportation logistics or the devices
themselves.
Digitizing the supply chain is innovative and inevitable. The predictive models, sensor (IoT) events, and business
rules for optimized supply chain execution are all automated and operationalized through dynamic case
management. Digitization spans the extended manufacturing digital enterprise, including OEMs, parts suppliers,
logistics, and transportation, to name a few.
Digitizing the Supply Chain

WWW.PEGA.COM13
Introducing Machine Learning Strategies to Provide Predictive Analysis
What innovations can be introduced on the shop floor?
Machine learning software uses computational statistics to learn and then predict outcomes based on
independent feedback entered into the model. Many online search engines use this self-learning approach,
which, in fact, is also a type or subset of artificial intelligence.
Today’s most reliable search engine models incorporate an underlying human behavior distribution model,
derived from years of data collected from the Internet. Take, for example, searches you’ve conducted on Google
or Amazon. How often did their algorithms predict what you were searching for?
MACHINE LEARNING ON THE SHOP FLOOR
We can apply the machine-learning strategies from these efficient search engine models to the manufacturing
shop floor.
We know that:
ƒƒ Machines are very good at performing tasks repetitively based on parameterized configuration
ƒƒ Machines are not as good at auto-adapting to special circumstances
ƒƒ Humans, on the other hand, are very good at adapting to special circumstances
To improve the adaptability of machines, we must model the human in such a way that machines will be able
to adapt to certain sets of circumstances. The auto-learning (auto-adapting) factory will use machine learning,
incorporating feedback from operators and engineers (human), as well as inter-machine information. The goal is
auto-predicting and prescribing actions around special circumstances.
USE CASE
In companies using hundreds and thousands of machines for production, it is often hard to know which machine
to work on or investigate first when processes start to go out of control.
Today we have an operator or engineer on each machine to start that investigation, but with machine learning,
incorporating user feedback (operator or engineer) can optimize the process by indicating which machines to
investigate first. The model could predict which one, based on parameters, SPC charts and previous engineer/
operator input.
The engineer or operator who accesses this new knowledge would in turn provide feedback so that the machine-
learning algorithm would reevaluate the model to make a more precise prediction in the future.
.11 Innovations on the Shop Floor

WWW.PEGA.COM14
Improved Processes Through Data-Based Next Best Action Predictions
If we envision the future of manufacturing, we can easily imagine a centralized cloud platform collecting machine data
such as Statistical Process Control (statistical models that can highlight when there are potential problems in a process or
machine) information and predicting which machines to investigate in the case of a suspect out-of-control process.
This future offers the “next best action” type of prediction for defects and defective products. It optimizes the lifetime value
of the machines without having to rely upon a significant number of operators or engineers.
NEXT BEST ACTION
Tying the next best action predictions (prescriptive) to information that not only comes from the shop floor but also
from the products produced (external) could have the following advantages:
ƒƒ Capturing data from the device about which components are potentially at the end of their useful life
ƒƒ Enabling the input of external data into the algorithm related to the functioning of the system components
ƒƒ Creating a new external feedback system to determine potential unforeseen out-of-control processes
Ultimately, employing next best action predictions creates a prescriptive process that is significantly more efficient
than one relying solely on feedback input from the engineer or operator.
USE CASE
Imagine that information was fed to robots that fix machines or to self-healing machines. Sensors providing information about
the lifespan of a component on the shop floor are augmented with return information of sensors in products for the sake of
sustainability.
Extending this vision further, as shop floors become more efficient with the use of external data for preventative maintenance
and sustainability motives, it is easy to see that the supply chain will become efficient as well. In this ultra-informed environment,
the supply chain can become a “bid-on-demand” model—a supply chain market similar to a commodities market.
Imagine a shop floor that bids when it predicts resource needs, identifying the best bidder with the fastest turnaround. Bids
are based on production, life of the machine, life of the product, life of the component, and sustainability requirements.
The whole shop floor is working in synchrony and quasi-automatically.
This is the future of manufacturing, and it is within reach now.
.12 The Shop Floor
Predictions:
• Which machine to investigate
next out of control
• Machine learning for
optimal investigation
(learn from humans)
Prescription:
• Next best action for
defects or defectives
• Increase value
lifetime of machines
and production line

WWW.PEGA.COM15
Specialized Manufacturing Strategies Providing Agility and Customer Input
NEW MODELS OF MANUFACTURING
Mega manufacturers need to keep in mind that trends like the following will enhance their business:
ƒƒ Micro manufacturing:[24]
The paradigm shift to smaller manufacturing plants producing customized products,
where the extreme micro-plant is the garage or basement of an individual
ƒƒ Social manufacturing:[25]
Linking multiple, smaller agile manufacturers to produce a larger aggregated product.
ƒƒ Mobile manufacturing:[26]
Mobile micro manufacturing able to move from one geographical location to another.
Internet innovation is unstoppable. One emerging trend we have witnessed over the last few years is social
manufacturing. Imagine living on a street with four houses in rainy Seattle. Three of the houses each have a 3D
printer (purchased at Home Depot for less than $1,000/each[27]
). The last house has a sewing machine.
The first house uses its 3D printer to produce the rounded handle of an umbrella, and the second house produces
the shaft. The third house produces the skeleton to the fabric, and the fourth sews fabric for this umbrella. Working
in tandem, they’ve produced a product that is invaluable at that location and can be customized quite easily.
Each house is an example of micro manufacturing as it focuses on a specific custom product built in small capacities
for a very specific need. If we imagine that the house is mobile and can be placed at the center of an event for
instant production of a product directly linked to the event, we now have an example of mobile manufacturing.
Although it may seem unconceivable today, in a few years this type of social manufacturing model may revolutionize
the way manufacturing is done.
Each one of these trends allows the manufacturer to become more agile and more in tune with constantly changing
needs. The manufacturer is closer to the consumer, and can be more innovative, changing the supply chain methods.
To negotiate this paradigm shift successfully, manufacturers must have the right tools.
.13 The End of Monolithic Manufacturing?
Predictions:
• Automatic Self-service
supply chain
• Micro and mobile factories
– directly attached to need
• Sentiment/text analysis to
predict optimal location
Prescription:
• Timely and
optimized location
• Adaptive needs
based on geography
and demographics
With the proliferation of 3D
printing, together with social
and additive manufacturing,
smaller innovative
manufacturing corporations
have the potential of
competing with the biggest
mega manufacturers…the
latter will have to adapt.

WWW.PEGA.COM16
Realizing the Power of all Assets in the Adaptive Digital Factory
Now we bring it all together—the execution and continuous optimization of the adaptive digital factory.
Information is generated from a diverse set of sources, such as suppliers, manufacturing units, machines and
operators, robots, products, consumers and sustainability systems. A system will be required to orchestrate it all.
Data from all of these sources is transformed so that it can be analyzed. This analysis in turn enables predictions
and self-learning outcomes.
DYNAMIC CASE MANAGEMENT IN ACTION
As indicated earlier, dynamic case management is the platform that aggregates, orchestrates, and facilitates
the execution of all the assets enabling the adaptive digital factory. These assets include process fragments,
predictive and self- or machine-learning adaptive models, as well as data or content from systems of records.
The dynamic cases are executed in the adaptive steady-state execution of the adaptive digital factory.
These cases orchestrate the entire lifecycle from production to execution[28]
—connecting the consumer to
manufacturers, services, and support.
Events in any phase of the production or servicing may trigger dynamic cases that will orchestrate the decisions,
sub-cases, and business processes to respond or handle the event. If machine learning is used in the predictions
or any type of adaptive predictions, it is conceivable that the actions, cases, and business processes actually may
be enhanced on the fly.
This process is adaptive, in terms of both orchestrated execution as well as continuous improvements in the
models or assets themselves (i.e., changes in processes, rules, decisions, etc.).
.14 Dynamic Case Orchestration
Sub-Cases Case
Content
Case
Events
Processes
(w)
Case Data Case
Business
Objectives
Decisions (w) Dynamic Case
Management
IoT Analytics & Machine Learning
Data)
Predictive and Adaptive
Orchestration
Machine &
Operator
Robots or
Self Fixing
Machine
Consumer Sustainability(Suppliers Intelligent
Product
Micro
Manufacturing
(w)

WWW.PEGA.COM17
For Adaptive Digital Factories
The dynamic case management platform aggregates, orchestrates, and facilitates the execution of all the assets
enabling the adaptive digital factory to build optimal connectivity on the shop floor and external connectivity with the
suppliers, Things, and the consumer. This adaptive process delivers the benefit of continuous improvements in the
models and assets, but it also requires careful consideration of the privacy and security issues related to connectivity.[29]
SECURITY THREATS: In any network, security breaches are a regular and serious concern. Hacking typically
focuses on stealing information or penetrating an environment to add malware to take control of it. Within the
world of Big Data, IoT, and the factory of the future, one needs to be conscious of how these and additional threats
impact the adaptive digital factory.
Consider the following examples:
ƒƒ A hack designed to control remote connected devices—either machines on the shop floor or the
manufactured connected product.
ƒƒ A hack that accesses and manipulates sensitive or private consumer data, creating malicious executions of
decisions, business processes, or dynamic cases.
ƒƒ A hack that implants malicious errors into sensitive data. Machine learning and adaptive predictions models
depend on good data to produce positive and realistic predictions. Any malicious data introduced into the
models will in effect create wrong predictions, potentially creating negative or even hazardous outcomes.
Clearly, security for Big Data, IoT, and factory of the future needs to be in the forefront of planning[30]
.
SECURITY PLANNING:Given the execution stack of the adaptive digital factory, you must address privacy
and security at all levels—from the lower or bottom edge devices all the way to the consumer interacting with the
connected products. Each of the following requires serious security consideration.
EDGE DEVICES: Connected products and smart machines on the shop floor need to be able to protect the
controls (and the information they sense or generate) from unauthorized access.
SHOP FLOORS: The vulnerabilities of increasingly automated controllers (e.g., those leveraging robotics) and the
systems of adaptive digital factory shop floors must be addressed. As operational robots and machines connect
to the information technologies and applications of the enterprise, the potential for privacy and security violations
will only increase.
THE CLOUD DATA AND CONTROLS: Increasingly, Adaptive Digital Factories rely on the cloud for storing product
and sensor data, as well as various types of controls for connected products. A variety of cloud security standards
and protection from public cloud to hybrid cloud and private clouds do exist, and should be evaluated.
BIG DATA/THING DATA: Of paramount concern is the privacy and security of data gathered and analyzed
throughout the entire lifecycle of production. Access control authorization, authentication for individual data entities
and aggregate data must be enacted together, using automated risk mitigation policies.
REMOTE CONTROL: Manufacturer or service providers occasionally control the connected products remotely,
typically through automated software updates and controls for maintenance. In either case, unauthorized access
checks must be embedded in the remote interactions with the connected products.
DYNAMIC (BUSINESS) CASES: Finally, dynamic cases orchestrate humans, Things, and enterprise applications, as
well as trading partners. For that reason, security, privacy authorization, and authentication with multiple levels
must be enacted for each category of participants (humans, robots, applications, etc.) in the end–to-end production
servicing dynamic cases.
.15 Security Concerns

WWW.PEGA.COM18
The Value of Top Down Digital Business Transformation
The products, parts, and suppliers
that make up Adaptive Digital
Factories may suffer from poor
data quality and inconsistencies,
which can lead to flawed and
even risky decisions. Master data
management(MDM)[31]
is essential
in modern manufacturing.
However, not all MDM
approaches efficiently manage
consistency across the value
chain.
MASTER DATA MANAGEMENT
The “bottom up” IT approach : Often organizations attempt to address the MDM issues through “big bang” master
data projects, tools, and even MDM centers of excellence for governance. The technology and data consistencies are
formidable and important to address, including data cleansing, addressing missing data, data consistency, ETL, and
data integration.
The danger is the enormous effort required to normalize master data without prioritizing through the business
objectives. One common problem is that initiatives attempting to resolve master data challenges often do so in a silo.
The “top down” digital business transformation approach: A more optimal approach is to treat MDM challenges
as part of overall continuous improvement initiatives, especially through end-to-end dynamic case management solutions
that connect silos that touch and manipulate master data. This “top down” approach prioritizes transformational projects
with MDM improvements and balances risk with business value.
These technical database issues must be addressed, but with a revamped approach to priorities. In the adaptive digital
factory, solutions to master data issues are now driven by a “Think Big…But Start Small” governance. Earn quick wins, build or
achieve the needed master data rigor, and then expand with additional transformational solutions that include MDM.
The core to this approach is a DCM-enabled layer that wraps and modernizes legacy systems. Iterative and agile
methodologies provide the adaptive case management solutions that are essential to the digital business transformation.
The MDM governance and implementation can be embedded in the DNA of the case automation methodologies and lead
to the discovery of data sources and the optimization needed to manage data within the DCM layer.
MASTER POLICY MANAGEMENT
Digital business transformation needs a top down approach for master data. It also needs a top down approach for
master policies.
Consider the scenario where your discount depends upon the channel you use to interact with the digital enterprise. If
you go to the Web, you get one discounted price, but if you call a customer sales representative, you get another! Both
channels might have the exact and precise information about you as customer or the product or service you are trying to
purchase (master data), but the pricing is defined by the channel.
This frustrating consumer purchase scenario clearly illustrates a lack of master policy, which would guarantee consistency
in the various phases of digital manufacturing. These decisioning policies apply to price in our example, but just as easily
could pertain to regulatory compliance, supplier policies, warranty policies, production, and quality.
.16 Master Data and Master Policy Management
Sub-Cases Case
Content
Tasks Case
Events
Processes
Case DataCollaboration Case
Stages
Case
Workers
Case
Business
Objectives
Decisions
Dynamic Case Management
1 2 3

WWW.PEGA.COM19
Transformative Digital End-to-End PLM[32]
Process
The Adaptive Digital Factory of the future will manage the product lifecycle more efficiently than ever by leveraging the
value from smart connected assets, such as machinery and assembly tools on the shop floor.[33]
Real-time data from
physical assets can be integrated with other systems, such as ERP, CRM and data warehouses, to provide in-depth
understanding of the production flow and quality.
TRANSFORMING THE PRODUCT LIFECYCLE
In the adaptive digital factory, IoT data from connected assets provides visibility into new opportunities to
improve daily operations. The following are some examples of those potential improvements:
ƒƒ By monitoring the usage and behavior of the equipment, the manufacturer can discover better ways to
perform maintenance. By measuring and analyzing critical data points, the factory can improve its overall
equipment efficiency.
ƒƒ Early detection and notification of issues can trigger proactive resolution of the problem and prevent costly
downtime.
ƒƒ When alerted to a potential issue, the engineers will also be informed of any critical information such as whether
the repair is covered under warranty or not.
ƒƒ The equipment manufacturer can use real-world usage data from the devices to enhance or upgrade features
and functionality in the planning and design of the next-generation of products.
ƒƒ Reacting dynamically to detected usage patterns and recurring issues will positively impact customer satisfaction
and loyalty.
ƒƒ Product development enhanced by using real-time data to improve and streamline product design, resulting in
faster time to market and increased competitive advantage.
The proliferation of low cost 3D printers has contributed to the acceleration of rapid prototyping and new product
introduction.
In the adaptive digital factory, new products are
developed in days, not months. Smart connected
devices improve the collaboration and flexibility
across the entire value chain from the supplier to
the factory to the end customer. Harnessing IoT data
allows the plant manager to adapt to changes more
quickly, whether it is a new product introduction, or
fluctuations in the supply of a particular component
or in customer demand.
.17 Evolved Product Lifecycle Management (PLM)
Supplier
Digital
Cross Functional
Interactions
Customer
Innovate Plan Engineer Operate Service
Market
& Sell
Conceptualize
Innovate
Design
Procure
Produce
Service
Warranties
Support

WWW.PEGA.COM20
Pega’s End-to-End Solution for Warranty Management
Device-Directed WarrantyTM[34]
augments Pega’s Manufacturing Application with connected devices. One of the
most common warranty claim scenarios aligns with the use cases discussed in Chapter .06.
The Process of Everything[16]
realizes the end-to-end warranty case processing with connected devices. More
specifically, an exception event instantiates a dynamic case that manages the end-to-to-end processes involving
Things, the OEM manufacturer, service technicians, potentially suppliers, the consumer, and, of course, the
distributor or the dealer.
The Adaptive Digital Factory®
leverages the powerful Pega platform with decision management in automated
dynamic cases, the Pega Warranty Management configurable IP, and the Pega IoT connectivity capabilities. More
specially, with Device-Directed WarrantyTM
manufacturers will gain the following capabilities:
ƒƒ IoT Diagnostics: Things will have on-board CPU and execution capabilities or will be able to connect to a
device that has on-board execution for the device or Thing via low power connectivity.
ƒƒ Automatic Updates of On-board Device Software: Manufactured devices (either edge devices or
gateways) often have sophisticated software that can be updated remotely by the manufacturer.
ƒƒ Automatic Sense and Data of Measures from Edge Device: A manufacturer will have the capability to
gather data from the device or ping it for specific measurements and analysis.
ƒƒ Automatic Control for Maintenance: Devices can be controlled remotely or through on board decisioning
software.
ƒƒ IoT Supply Chain and Parts Return: The edge device that needs to replace a defective device can
be monitored at any point in the supply chain—from the supplier to the distributor or customer. In this
scenario, a defective device that needs to be replaced can be monitored to confirm its return to the supplier
or the OEM manufacturer.
ƒƒ IoT Repair or Parts Validation: Once a repair or replacement is completed, the manufacturer can validate
the fix and its compliance to the manufacturing management policies and procedures.
.18 Device-Directed WarrantyTM
Supplier Manufacturer Logistics Service

WWW.PEGA.COM21
The Power of End-to-End Digital Automation of Maintenance Processes
The Adaptive Digital Factory, as we have described it, goes well beyond reactive recalls and repairs, and offers
improvements even beyond the Lean and Six Sigma methodologies that have improved operations for decades.
Lean manufacturers have long subscribed to the idea of total productive maintenance (TPM), which has the goal of
improving overall equipment effectiveness as measured by performance x availability x quality. With TPM, techniques
such as quick changeovers are used to increase productivity and reduce downtime.
In the Adaptive Digital Factory, digitization of processes and decision management are used to avoid the
consequences of equipment failure. The idea is to prevent the failure before it occurs by replacing or repairing worn
parts. In the current environment with IoT and Big Data, there is an opportunity to modernize maintenance proactively.
PREDICTIVE MAINTENANCE
The advent of predictive maintenance[4]
represents an important milestone in the digital evolution of increasingly
automated and intelligent maintenance.
With predictive maintenance, Big Data analytics of messages or events emanating from Things are analyzed to identify
what could be done for maintenance. Data is available from sensors that track operating conditions, how equipment is
being used, and even how and why breakdowns occur. These insights reduce waste and create a leaner organization.
Consider the example of a large farm equipment manufacturer using new technology to help farmers improve
agricultural yields. The manufacturer has access to an unprecedented amount of performance data from connected
devices.
The manufacturer’s goal is to use data analytics to predict the point in the future when maintenance will be required.
The IoT allows devices to collect data about the equipment and its condition, triggering an alert when certain
thresholds are reached or an anomaly occurs, and providing recommended actions to take next.
PRESCRIPTIVE MAINTENANCE
By leveraging end-to-end digital automation of maintenance processes, as well as decisioning (self-learning or
adaptive), the transformative optimization of maintenance is achieved through prescriptive maintenance.
With prescriptive maintenance, the objective is what maintenance should be done in an IoT digital environment.
When paired with IoT, prescriptive maintenance[35]
is holistic. It encompasses the following:
ƒƒ Total Productive Maintenance:[36]
Continuously improving overall equipment/device effectiveness.
ƒƒ Predictive Maintenance: To analyze events and data streams especially from IoT Big Data sources and
identify what could be done.
ƒƒ Self-Learning and Adaptive Maintenance: Predictive analytics looks at past data and attempts to mine
and discover models. A complimentary approach is to continuously monitor and learn from the behavior
of IoT data and adapt maintenance priorities accordingly. Remote devices can track the machine’s location,
fuel consumption, and utilization. Data is analyzed to perform initial diagnostics and prevent problems
from occurring. The end result is improved uptime and increased machine and operator productivity. New
technology and predictive and adaptive analytics help ensure the future is leaner than ever before.
ƒƒ Knowledge Work and Policies: The knowledge of the experts is captured in decision rules and involved
throughout the maintenance lifecycle.
ƒƒ Automation of the Value Stream: The automated value stream orchestrates end-to-end the connected
device (IoT), supplier, manufacturer, and consumer. This value stream ties the consumer’s device to the
entire lifecycle of maintenance via automated and digitized processes, with complete visibility and control of
the end-to-end processes.
.19 Digital Prescriptive Maintenance

WWW.PEGA.COM22
Omni-Channel to Omni-Device
The Internet of Things is also transforming the alignment of customers to the manufacturers.
Omni-device[37]
is really a specialization of omni-channel that leverages Internet of Things. Typically, customers
interact with digital enterprise across many channels:
Omni-device: With the rise of Things, devices in our homes (e.g., appliances), connected vehicles, workplaces,
or connected wearables are becoming channels. What does omni-device mean and how does it manifest itself
with IoT? There are four essential omni-device capabilities that provide significant potential in optimizing and
transforming the customer experience through IoT:
1. Customer-Device Interaction: Manufacturing and High Tech companies are realizing the tremendous
potential of smart connected devices for applications in connected homes, vehicles or industrial applications.
GE for instance is adding if-then smarts and intelligence to connected appliances to allow the consumer the
ability to indicate what they want to do with the appliance (e.g., turn it off after 20 minutes, inform via text
message if the oven is heated, etc). The connected devices will of course provide both sales and marketing
opportunities.
2. Things as CRM Channels: IoT devices are also becoming channels. Just as a mobile device or a browser
handles customer service, Internet of Things or devices can also be used to promote products, services,
cross-sell, or upsell. For instance, automobiles connected through 4G LTE connections can offer Wi-Fi as well
as vehicle diagnostics and maintenance opportunities. However, this connectivity also provides opportunities
for advertising, sales, and marketing – all convenient for the consumer
3. Connecting With Manufacturers: The third use case for omni-device deals with connecting the consumer
with the manufacturer of the devices. With the emergence of connected products, manufacturers are now
much closer to the consumers. Sensors are providing incredible feedback on how the device is being used
as well as the preferences of the consumer. There are, of course, healthy discussions and concerns around
privacy and security in this instance. Nevertheless, this connectivity and continuous sensor feedback is in fact
transforming the relationship between the manufacturer and the consumer.
4. Extending the Manufacturers’ Ecosystem to Optimize the Customer Experience: The consumer is
constantly on the move and connected through their devices. This includes their offices, cars, and homes.
Now, due to this connectivity Manufacturers are establishing new innovative relationships with other
merchandising channels to conveniently cross-sell or up-sell to the customer. For instance, OnStar now
offers AtYourService with various offers to the consumer (e.g. coupons for coffee, gas, restaurants, even
retail) that conceivably takes into consideration the geolocation as well as the context of the customer.
.20 Aligning Customers and Manufacturers
Web Email Chat Social Mobile Sensors

WWW.PEGA.COM23
The Impact of the Cyber-Physical Ecosystem
Even though connected devices on the Internet have been around since its inception, only recently has the Internet
of Things had a significant impact on industries, especially manufacturing. From connected homes, cars, cities to
industries, such as healthcare, insurance, manufacturing, and utilities, connected and increasingly intelligent devices
are transforming entire ecosystems. The IoT sectors are often characterized or contrasted between Consumer IoT
and Industrial IoT.
The connected (or “smart”) home is the quintessential example of the former. Connected industries, such as
process or discrete manufacturing, are the examples of the latter. Things could be small devices such as smart
watches or health monitoring devises, that communicate to the Internet via smartphones (and increasingly the
smart phone itself is providing sophisticated monitoring functions). The other end of the spectrum are huge
machines, such as turbo engines, wind energy farms, oil wells, manufacturing robots, that are increasingly intelligent,
semi-autonomous, or even targeting complete autonomy.
In fact, the impact of IoT on industrial ecosystems has been so great that the emergence of devices with robust physical-
cyber connectivity, often called the fourth industrial revolution (aka Industry 4.0[39]
). Frost & Sullivan, the global research
and consulting organization, dubbed the integration of these physical devices into the software systems “The Industrial
Internet.”[40]
Frost & Sullivan sees the Industrial Internet (spanning the spectrum of machine-to-machine communication,
people, software, and Big Data analytics) as the means to promote innovation, improve industrial machine maintenance,
as well as operational processes.
This new connected ecosystem encompasses physical machines, devices, network connectivity, and various types of
computing (process, operations, and analytics)—the short list of key elements for optimized asset management and
business outcomes.
STANDARDIZATION AND INDUSTRY ALLIANCES FOR IOT
A number of consortia, initiatives, references architectures, and standardization efforts have arisen to address
the wide spectrum of levels from the physical to reference architectures[41]
. Robust interoperability within and
across industries as well as standardization of protocols of IoT architecture stacks are necessary for successful
IoT deployments. Standards for the IoT stack include ZigBee, Z-Wave, BlueTooth LE,6LowPAN, Thread,WiFi,
Cellular and others. Most of these standards support the 128-bit IPv6[42]
addressing protocol. IPv6 will support
340 trillion, trillion, trillion IP addresses. In other words, almost an infinite number of Internet addressable devices
and services. Industry solution and service providers in the IoT space have started to collaborate on a number of
consortia that often span standards, interconnection technologies, reference architectures and best practices.
Industry organizations include AllSeen Alliance, Open Interconnect Consortium, Industrial Internet Consortium,
and Industry 4.0.[43]
Let’s look at the last two groups to get a sense of their mission.
ƒƒ Industrial Internet: In 2014, AT&T, Cisco, General Electric, IBM, and Intel founded the Industrial Internet
Consortium[44]
. This consortium is “setting the architectural framework and direction for the Industrial
Internet.” Among many initiatives it has defined an overall reference architecture for the Industrial Internet
application of IoT, encompassing all related phases, technologies, and solutions. This RA provides guidance
covering all functional domains in industries leveraging IoT, including business, operations, applications,
and especially controls involving connected devices. Manufacturing is, of course, a major sector, but other
verticals, such as healthcare, public sector, transportation, and energy, are addressed as well. The scope of
the RA and the IIC overall is to provide architectural best practices in combing, aggregating and connecting
people, processes, data, and IoT.
.21 Industry 4.0 and the Industrial Internet

WWW.PEGA.COM24
ƒƒ Industry 4.0 (aka industrie 4.0): Industry 4.0 is a key high-tech manufacturing initiative introduced in 2011
by the German Federal Government. Industry 4.0 “is a term applied to a group of rapid transformations in
the design, manufacture, operation and service of manufacturing systems and products.”[45]
This initiative
encourages manufacturers of all sizes to achieve efficiencies in maintenance, operations, and energy savings.
Cyber-physical systems with sensors are among the key digital technologies that are leveraged in Industy 4.0.
Another major focus of Industry 4.0 is mining the data or information generated by the assets and applying
real-time analytics to predict, for instance, failures in maintaining assets. With connected devices and assets,
Industry 4.0 envisions active monitoring of data generated by the entire supply and value chain as a way of
managing logistics and enacting real-time controls.

WWW.PEGA.COM25
As we have seen, digitization means disruptive megatrends in connectivity, social, Internet of Things, Big Data
analytics, cloud, and mobility. The transformational aspect of digital technologies is paving the way for digital
innovations that will change the landscape of manufacturing.
WHY DIGITAL TRANSFORMATION IN MANUFACTURING?
As digitization increases at a furious pace, so too does the pace of change for products and services. Yet, today’s
economy means “doing more with less,” so companies face challenges in keeping up with demands from each link
in the supply chain, through to the consumer.
Within these companies, IT is challenged by the disruptive impact of IoT technologies and the needs of more tech
savvy business stakeholders. The IoT is transforming the entire end-to-end manufacturing value stream.
Companies must not underestimate the impact of aligning the cyber world with physical connected devices. These
devices generate enormous amounts of information that must be mined, analyzed, and acted upon. With the
proliferation of connected devices, traditional manufacturing models no longer apply.
Historically, the physical Things and software in the realm of IT are in siloes. In the transformed manufacturing
model, the traditional information technology model is augmented with operational technology (OT). Embedded
software solutions with physical devices and the maintenance and control of Things is the realm of OT. The physical
and virtual worlds are coalescing in the new digital enterprise.
Embedded sensors, software, controllers, and connectivity are creating a digital revolution in manufacturing and
aftermarket services. Breakthroughs in networking, edge computing, cloud technology, energy efficiency, and
miniaturization are converging to create low-cost processing power and data storage everywhere. Computers in
machines, gadgets, and wearable devices are streaming data about their operations and conditions.
Things are already generating more data than people or applications. Translating that data into insights and
intelligent decisions is the key to effective analytics. Intelligent, future-facing decisions can be derived from Thing
Data, provided the company applies a digitized process context to orchestrate tasks for actions executed by
humans, enterprise applications, supply chain partners, and increasingly smart Things or devices. The next industrial
revolution is already here.
.22 The Digital Transformation of Manufacturing

WWW.PEGA.COM26
References
[1] Khoshafian, S. (2014a). Rise of Things: IoT’s role in Business Processes.
http://www.informationweek.com/mobile/mobile-devices/rise-of-thingsiots-role-in-business-processes/a/d-id/1317010
[2] http://searchmanufacturingerp.techtarget.com/definition/Industrial-Internet-of-Things-IIoT
[3] Khoshafian, S. (2014b). Dynamic Case Management for the Modern Worker.
http://www.pega.com/insights/articles/trend-6-dynamic-case-management-modern-worker
[4] Levitt, J. (2011). Complete Guide to Predictive and Preventive Maintenance. New York: Industrial Press Inc.
[5] For a description of Big Data and Fast Data: http://en.wikipedia.org/wiki/Big_data
and http://www.infoworld.com/article/2608040/big-data/fast-data--the-next-step-after-big-data.html
[6] http://3dprinting.com/what-is-3d-printing/
[7] See http://www.pega.com/insights/articles/trend-10-adaptive-digital-enterprise-part-1-gazing-crystal-ball-top-10-
trends-2014 and http://www.pega.com/insights/articles/trend-10-adaptive-digital-enterprise-part-2-gazing-crystal-
ball-top-10-trends-2014
[8] http://ioeassessment.cisco.com/
[9] Khoshafian, S. and Rostetter, C. (2015a). Digital Transformation in Manufacturing. In CIO Review – page 42-43
http://magazine.cioreview.com/June-2015/Manufacturing/
[10] Khoshafian, S. (2015a). Digital Transformation of CRM through Internet of Things (IoT).
http://www.pega.com/insights/articles/digital-transformation-crm-through-internet-things-iot
[11] https://en.wikipedia.org/wiki/Internet_of_Things
[12] Khoshafian, S. (2014c). Intelligent BPM: The Next Wave for Customer-Centric Business Applications. Pega eBook
publication. http://www.pega.com/insights/resources/intelligent-bpm-next-wave-customer-centric-business-applications
[13] Khoshafian, S. (2015b). Internet of Things: Achieving Business Objectives through Dynamic Case Management.
http://www.siliconindia.com/magazine-articles-in/Internet_of_Things_Achieving_Business_Objectives_through_Dynamic_Case_Management-BBSY475164515.html
[14] Atos (2012). The convergence of IT and Operational Technology. White paper from Atos
https://atos.net/content/dam/global/ascent-whitepapers/ascent-whitepaper-the-convergence-of-it-and-operational-technology.pdf
[15] https://en.wikipedia.org/wiki/ANSI/ISA-95
[16] Khoshafian, S. & Schuerman, D. (2013). Process of Everything. Published in iBPMS: Intelligent BPM Systems,
Foreword Jim Sinur, edited by Layna Fischer. Lighthouse Point, FL: Future Strategies, Inc., Book Division.
http://www.futstrat.com/books/iBPMS_Handbook.php
[17] ] Khoshafian, S. & Adams, K. (2015). Internet of Things: A New World of Analytics, Digital Innovation and
Transformation. PegaWORLD 2015 Breakout Session https://www.pega.com/insights/resources/pegaworld-2015-
internet-things-new-world-analytics-digital-innovation-and-0
[18] https://en.wikipedia.org/wiki/OODA_loop
[19] Walker, R. (2015). For Smarter Customer Engagement, Think Like a Fighter Pilot.
http://www.pega.com/insights/articles/smart-customer-engagement-think-fighter-pilot
[20] https://hadoop.apache.org/
[21] Brynjolfsson, E. & McAfee, A. (2014). The Second Machine Age. New York: W. W. Norton & Company Ltd.

WWW.PEGA.COM27
[22] Hugos, M. H. (2011). Essentials of Supply Chain Management, Third Edition. Hoboken, N.J.: John Wiley & Sons.
[23] Council, A. & Petch, M. (2014). 3D Printing: Rise of the Third Industrial Revolution. Published by Gyges 3D.
[24] Wall, M. (2013). Make-It-Yourself: The rise of the micro-manufacturers.
http://www.bbc.com/news/business-24203938
[25] The Economist: Special Edition on Manufacturing and innovation (2012). Collaborative Manufacturing: All Together
Now. http://www.economist.com/node/21552902
[26] Columbus, L. (2015). 10 Ways Mobility Is Revolutionizing Manufacturing.
http://www.forbes.com/sites/louiscolumbus/2015/04/20/10-ways-mobility-is-revolutionizing-manufacturing/
[27] http://www.homedepot.com/b/Tools-Hardware-Power-Tools-3D-Printers-Accessories/N-5yc1vZc7mr
[28] Khoshafian, S. (2015c). Digital Transformation: Digitizing the Value Stream.
https://www.linkedin.com/pulse/digital-transformation-digitizing-value-stream-setrag-khoshafian
[29] Bergin, B. (2015). Industrial IoT: Security Solutions. http://www.manufacturing.net/news/2015/10/industrial-iot-
security-solutions
[30] Federal Trade Commission Report (2015). Internet of Things: Privacy and Security in a Connected World. FTC Staff
Report. https://www.ftc.gov/system/files/documents/reports/federal-trade-commission-staff-report-november-
2013-workshop-entitled-internet-things-privacy/150127iotrpt.pdf
[31] Khoshafian, S. (2014d). Digital Transformation through DCM for Master Data & Policy Management.
http://www.pega.com/insights/articles/master-data-management-adaptive-digital-enterprises
[32] http://www.product-lifecycle-management.info/
[33] Khoshafian, S. & Davison, M. (2011). Discover the 7 Strategic Advantages of BPM for Product Lifecycle Management.
http://www.pega.com/insights/resources/discover-7-strategic-advantages-bpm-product-lifecycle-management
[34] Khoshafian, S. (2015d). Device Directed Warranty.™ Pega eBook publication.
http://www.pega.com/insights/resources/device-directed-warrantytm/
[35] Khoshafian, S., and Rostetter, C. (2015b). Digital Prescriptive Maintenance. In Internet of Things, Process of
Everything, BPM Everywhere edited by Layna Fischer. http://futstrat.com/BPMEverywhere.htm Lighthouse Point,
Florida: Future Strategies, Inc., Book Division.
[36] http://en.wikipedia.org/wiki/Total_productive_maintenance
[37] Khoshafian, S. (2015e). Omni-Channel to Omni-Device: Evolved CRM with IoT.
http://www.pega.com/insights/articles/omni-channel-omni-device-evolved-crm-iot
[38] http://www.pega.com/insights/resources/build-change-omni-channel-ux
[39] https://en.wikipedia.org/wiki/Industry_4.0
[40] See Industrial Internet research in http://ww2.frost.com/
[41] See http://www.iiconsortium.org/IIRA.htm and http://www.iot-a.eu/public
[42] IPv6 fact sheet http://www.internetsociety.org/deploy360/ipv6/factsheet/
[43] The Internet Society (2015). The Internet of Things: An Overview.
https://www.internetsociety.org/sites/default/files/ISOC-IoT-Overview-20151014_0.pdf
[44] http://www.iiconsortium.org/
[45] European Parliament (2015). Industry 4.0: Digitalisation for productivity and growth.
http://www.europarl.europa.eu/RegData/etudes/BRIE/2015/568337/EPRS_BRI(2015)568337_EN.pdf

WWW.PEGA.COM28
ABOUT THE AUTHORS
Dr. Setrag Khoshafian
VP,Chief of Strategy and Lead Evangelist, Pegasystems
One of the industry’s pioneers and recognized experts in Digital Enterprises, especially Digital Transformation
through IoT, Evolved CRM and intelligent BPM. Lead author of more than 10 books and more than 50 publications
in various industry and academic journals.
Carolyn Rostetter
Sr. Director, Manufacturing Industry Principal, Pegasystems
Business Optimization Leader in some of the world’s most respected organizations. As a Lean Six Sigma
Master Black Belt, provides vision and experience in areas such as Strategic Planning, Productivity, Quality,
Organizational Transformation, Change Management, Client Services, Enterprise Rationalization, and
Master Data Management.
Kenty Adams
Enterprise Architect: Data Science | IoT | Advanced Analytics | Jabil, Inc.
Strategist for programs of data science, advanced analytics and IoT analysis including application rationalization
and integration. Evangelist for neural networks, machine learning, statistical modeling, business/econometric
modeling, complex event management and the collaborative data science practice.

ABOUT PEGASYSTEMS
Pegasystems develops strategic applications for sales, marketing, service and operations. Pega’s
applications streamline critical business operations, connect enterprises to their customers seamlessly
in real-time across channels, and adapt to meet rapidly changing requirements. Pega’s Global 500
customers include the world’s largest and most sophisticated enterprises. Pega’s applications, available
in the cloud or on-premises, are built on its unified Pega 7 platform, which uses visual tools to easily
extend and change applications to meet clients’ strategic business needs. Pega’s clients report that Pega
gives them the fastest time to value, extremely rapid deployment, efficient re-use and global scale.
For more information, please visit us at WWW.PEGA.COM
© 2016 Pegasystems, Inc. All rights reserved. All trademarks are the property of their respective owners.

adaptive-digital-factory-whitepaper

More Related Content

What's hot (20)

Viewers also liked (12)

Similar to adaptive-digital-factory-whitepaper (20)

adaptive-digital-factory-whitepaper