iiot and big data analytics: architecture is being ... · pdf filetrial companies as well. ......
TRANSCRIPT
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
IIoT AND BIG DATA ANALYTICS: How Manufacturing System Architecture Is Being Transformed
lnsresearch.com
IIoT AND BIG DATA ANALYTICS:How Manufacturing System Architecture Is Being Transformed
lnsresearch.com
TABLE OF CONTENTS
Section 1: IIoT: State of the Market . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Section 2: Understanding Digital Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Section 3: Adoption of IIoT Connectivity and Big Data Analytics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Section 4: Building the Business Case and Recommended Actions . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
Introduction
The Industrial Internet of Things (IIoT) is a general term denoting the
concept that standard Internet technologies are broadly applicable
and will transform all areas of the industrial sector. Because of its
expansive and transformative nature, the concept and term “IIoT” is
starting to be used almost everywhere: manufacturers, automation
vendors, enterprise application vendors, system integrators, man-
agement consultants, government sponsored consortia, and indus-
trial companies as well.
As an analyst firm, LNS Research is dedicated to helping all of these
market players simplify, improve understanding, and more quickly
capture the value of emerging technology. In early 2014 LNS began
researching and advising companies on the IIoT, culminating in the
2015 research report: “Smart Connected Operations: Capturing the
Business Value of the IoT.” In this work we gave the first, now broadly
accepted, definition of the IIoT Platform, consisting of four main
buckets of capabilities: Connectivity, Cloud, Big Data Analytics, and
Application Development. We also put forth concepts of how this
platform would enable new data and system architecture that would
flatten existing hierarchies, provide data from anywhere to anywhere
capabilities, and enable next-generation business applications.
Since that time there have been dramatic moves by new start-ups
and many incumbent vendors, including announcements of home-
grown capabilities as well as mergers and acquisitions. All of these
moves have confirmed the assertion that for the foreseeable future,
the IIoT Platform space will be an ecosystem play that brings togeth-
er both IT and OT vendors to enable new business models.
In this new research, we will explore new survey data showing
the increased market adoption of IIoT Platform capabilities and how
these new technologies are transforming architectures today; not
some unknown date in the future. We will also examine how the LNS
Research Digital Transformation Framework can help industrial com-
panies overcome IIoT challenges and get started now on the journey
towards using Big Data Analytics to achieve a competitive advantage
and better business results.
5
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
Research Demographics
The data presented in this eBook represents over 300 completed
surveys and was collected from the middle of 2015 to the middle
of 2016. LNS Research deploys a social research model where our
online format English language surveys are open to the general
public. Companies participate in LNS Research surveys to gain
access to the LNS Research library, meaning survey participants
are research consumers as well. Each respondent is contacted with
multiple emails and phone calls and each response is reviewed by
an LNS Research analyst for accuracy.
The industry demographics of the survey largely match the
broader demographics of the industrial landscape, with discrete
being the largest segment, followed by process and batch indus-
tries. Our research also has a broad split across industries and
company sizes.
COLOR BY INDUSTRYCOLOR BY HQ LOCATION
Process Manufacturing
Discrete Manufacturing
Batch Manufacturing
North America
Europe
Asia/Pacific
Rest of World
2016 Metrics That Matter SurveyINDUSTRY
2016 Metrics That Matter SurveyREVENUE
2016 Metrics That Matter SurveyGEOGRAPHY
COLOR BY COMPANY REVENUE
Small: Less than $250 Million
Medium: $250 Million - $1 Billion
Large: More than $1 Billion
45% 49%41%
10%28%
15%
12%
37%
48%
15%
6
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
IIoT Adoption Platform Technology Adoption
A major change in the market from 2015 to 2016 was in regard to one
of the biggest challenges limiting the adoption of IIoT technologies.
In 2015, nearly half (44%) of companies did not know or understand
the IIoT. In 2016 this number has reduced to 19%.
Call it hype or call it hard work by many of the thought leaders
in the space, but the majority of companies now understand what
the IIoT is. In our 2015 research we showed that the IIoT market
was largely still an early adopter market but would likely follow the
typical adoption curve for major new technology innovations. We
also stated three things would have to occur to move the market
toward mainstream adoption:
• Time would need to pass
• The market would have to better understand
the technology
• Early adopters would have to prove results for
easier business case validation
Two of the three have occurred and it is now time for the market
to prove the value and demonstrate the business case, which we
will hopefully begin later in this eBook.
Please indicate how the IoT is impacting your business today
Do not understand or know about IoT
We understand and our customer demands
are driving us
We are still investi-gating the impact
We understand/are aware and see value to our oper-
ators/customers or both
We understand but see no impact at this time
We understand and have already seen
dramatic impact
0% 5% 10% 15% 20% 25% 30% 35% 40% 45%
2016
2015
19%
33%
18%
13%
8%
8%
44%
21%
16%
9%
6%
4%
2015: 44% 2016: 19%
Do not understand IoT:
7
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
Top IIoT Challenges
Unfortunately, building the business case and securing the funding
are the two top challenges facing IIoT technology adoption. Surpris-
ing to many, security concerns and technology scalability do not top
the list. This result is likely due to some companies just looking to
get pilot projects off the ground and not having tackled these tech-
nology issues yet. For other companies it is likely because they have
already done the research and are confident that IoT technology is
scalable to the industrial space.
In either case it is clear that building an effective business case is
the key to unlocking the potential of IIoT technology and Big Data
Analytics. In the next several sections we will show how building the
business case fits within the larger context of a Digital Transformation
Framework, how it must support these objectives, and how it needs
to be viewed as a journey rather than a singular, one-off decision.
Funding
Building a business case
Understanding what IIoT is and how it applies to your business
Security
Standards
Finding the right technology partner(s)
Gaining insight from Big Data
Developing new IIoT software applications
Company culture
Data gathering from legacy systems
Product design and development complexity
Hiring the right talent
Executive support
Scaling to 1,000s or 1,000,000s of devices
0% 5% 10% 15% 20% 25% 30% 35%
32%30%
26%25%
22%17%
16%14%14%
13%12%
8%8%
5%
What are the top challenges your company faces in deploying IIoT technology?(N=269, all respondents)
9
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
CEO/COO
Business Leaders
CDO, IT/OT Leaders
Functional Managers, SMEs
Business, IT/OT Practitioners
Digital Transformation Framework
Many industrial companies today are pursuing Digital Transformation
initiatives. What many organizations are missing is a systematic
approach to manage this transformation across all levels and functions
of the organization. The LNS Research Digital Transformation
Framework is designed to help industrial companies understand how
to connect all of these simultaneous and interconnected initiatives.
SOLUTION SELECTION
BUSINESS CASE DEVELOPMENT
OPERATIONAL ARCHITECTURE
OPERATIONAL EXCELLENCE
STRATEGIC OBJECTIVES
Eliminating Bias and Finding Long Term Partners
Evaluation
Team
Research
Pilot
RFPDISCOVERY
PLANNINGBUSINESS CASE
SELECTION
ProjectCharter
Defining Immediateand Long Term ROI
Managing IT-OT Convergence and Next-Gen IIoT Technology
Realigning People,Process, andTechnology
Reimagining BusinessProcess and Service Delivery
COSTS TOTAL YEAR 1 YEAR 2 YEAR 3 YEAR 4 YEAR 5
HARDWARE
SOFTWARE LICENSING
THIRD PARTY SOFTWARE
APPLICATION SOFTWARE
DOCUMENTATION & TRAINING
MAINTENANCE
INSTALLATION
INTEGRATION
LEGACY DATA LOADING
PROJECT MANAGEMENT
SUPPORT
TOTAL:
CONNECTIVITY
SMART CONNECTED ENTERPRISE
APPLICATIONDEVELOPMENT
CLOUD
BIG DATA ANALYTICS
IoT Enabled Business SystemsL4
Smart Connected Operations - IIoT Enabled Production, Quality, Inventory, MaintenanceL3
L2 L1 L0
IIoT EnabledNext-Gen Systems
L5 IoT Enabled Governance and Planning Systems
Smart Connected Assets -
IIoT Enabled Sensors, Instrumentation, Controls, Assets, and Materials
APMEHS
ENERGY QUALITY OPERATIONS
People – Process – TechnologyOperational Excellence Platform
OPERATIONAL EXCELLENCE SUPPORT
Fall short on any pillar and your OpEx platform becomes tippy
Fall short on two or more pillars and yourOpEx platform becomes totally unstable
DIGITAL TRANSFORMATION FRAMEWORK
The LNS Research Digital Transformation Framework offers a systematic approach
to undertaking simultaneous and interconnected IIoT initiatives
10
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
Digital Transformation Framework (Cont.)
STRATEGIC OBJECTIVES: At the highest level industrial companies today need to be thinking
about how many of these new technologies, like the IIoT, can
disrupt and transform products, value chain business processes,
and connected service delivery. At the strategic level, companies
should be doing 5, 10, and even 20-year planning, and often
these transformative visions are built around the competitive
differentiators of the firm, the changing nature of service delivery,
and existing models like Industrie 4.0, Smart Manufacturing, or
Smart Connected Assets.
OPERATIONAL EXCELLENCE: People, processes, and technology are the underpinnings of
Operational Excellence initiatives, and these are typically owned
by the business leaders in the organization. Leading companies
today have developed maturity models to help set goals and
growth plans for people, process, and technology capabilities
along with metrics programs to evaluate performance across
all areas of operations. Most companies have had Operational
Excellence initiatives in some form or fashion for 10 years or more.
Often these initiatives also incorporate the multiple management
systems and continuous improvement capabilities of the firm, like
Lean or Six Sigma. Moving forward, manufacturing companies
need to continue to evolve Operational Excellence initiatives to
not only be the continuous improvement engine of the company
but also the driving force for innovation. Often this means moving
to more of more of a lean, start-up mentality of “fail often and
fail fast,” with pilot projects that have the potential of delivering
much more than the typical 1%-2% benefits promised by most
continuous improvement initiatives.
OPERATIONAL ARCHITECTURE: Traditionally Enterprise Architecture has been owned by the IT
organization and has been typically focused on establishing robust
processes for evolving the enterprise application landscape and
supporting IT stack. Separately, automation, corporate engineering,
and/or advanced manufacturing (often now referred to as OT)
owned the technology architecture for plant level technology.
With the emergence of IIoT, LNS Research recommends industrial
companies adopt an Operational Architecture approach that
applies the formalized rigor and process of Enterprise Architecture
to the entire IT-OT stack. For this to be effectively accomplished,
industrial companies need to create supporting and collaborative
groups that incorporate both IT and OT, and as the role of Chief
Digital Officer emerges, the success of this new collaboration as a
key part of his or her charter.
BUSINESS CASE DEVELOPMENT: Often industrial companies begin business case development and
solution selection without also thinking about the connection
to broader Strategic Objectives, Operational Excellence, and
Operational Architecture. Typically these business case development
initiatives are successful when driven by deep subject matter experts
that understand both the process and technology.
11
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
Digital Transformation Framework (Cont.)
Identifying these experts can be a challenge, but often they are
located in advanced manufacturing, hybrid IT/OT roles, are a
leader within specific business functions, or are a technical fellow
supporting the organization. Although these other areas of Digital
Transformation do not need to be complete before a business case
is started, they are interconnected. As such, it is important industrial
companies do not view technology investments as a one-off business
case but rather as a business case journey that aligns with Operational
Architecture goals, depends on increasing Operational Excellence
maturity, and supports long-term Strategic Objectives. It is also
important to note that a strong business case will also incorporate
risk-based principles into the decision making and explicitly look at
‘no decision’ as an active choice.
SOLUTION SELECTION: Often industrial companies view Digital Transformation upside
down, starting with solution selection, which then drives all
other portions of the framework, rather than vice versa. Again,
with solution selection, it is important to put the activities within
the context of the broader initiatives. Solution selection is never
successful in a vacuum and when it is done in such a fashion, change
management becomes an insurmountable challenge and adoption
wanes. For success, build an effective solution selection process
that is quantitative to eliminate bias and a team that incorporates all
relevant portions of the organization, including IT, OT, and cross-
functional business leaders.
Solution selection should always be viewed within the broader context of
Digital Transformation initiatives; never in a vacuum
12
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
A New Model for Operational Architecture
In moving to a new model of Operational Architecture, industrial
organizations need to move to an expanded scope of Enterprise
Architecture. This expanded scope should account for managing
“things” with edge analytics and applications across the value chain of
suppliers, internal operations, customers, and products. It should also
span an application and analytics environment that includes cloud/
on-premise and time series/structured/unstructured data types. Upon
careful inspection this expanded model should also incorporate the
main components of the IIoT Platform: Connectivity, Cloud, Big Data
Analytics, and Application Development.
This expanded scope is also too broad to address the right level of
detail for making meaningful architectural decisions across the en-
terprise. LNS Research recommends a three-level approach, where
at Level 1 the entire scope is encompassed.
LEVEL 1OPERATIONAL ARCHITECTURE
Big Data Analytics, Collaboration, and Mash-Up Apps
Connectivity and Data Model
ANALYTICS & APPSANALYTICS & APPSANALYTICS & APPS
SUPPLIERS OPERATIONS CUSTOMERS & PRODUCTS
EDGE ANALYTICS AND APPLICATIONS
EDGE ANALYTICS AND APPLICATIONS
EDGE ANALYTICS AND APPLICATIONS
EDGE ANALYTICS AND APPLICATIONS
EDGE ANALYTICS AND APPLICATIONS
EDGE ANALYTICS AND APPLICATIONS
EDGE ANALYTICS AND APPLICATIONS
EDGE ANALYTICS AND APPLICATIONS
EDGE ANALYTICS AND APPLICATIONS
Applying an expanded scope to Operational
Architecture that includes the IIoT
Platform allows for the management of “things” across the value chain
13
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
A New Model for Operational Architecture (Cont.)
At the next level of detail, a particular element of the high level archi-
tecture should be driven into. For example, an organization’s Level 2
Operational Architecture for structured data analytics and apps would
largely map to the traditional scope of enterprise applications.
When building this architecture, LNS Research recommends not
focusing on the traditional applications, such as ERP, PLM, MES, SCM,
and CRM, but instead on the functional areas and map these to the
corporate systems/management systems/value chain systems used
across execution/planning/analytics. Then the different applications
can be mapped to this model, not vice versa.
LEVEL 2OPERATIONAL ARCHITECTURE
Structured Data Applications and AnalyticsTRADITIONAL ENTERPRISE ARCHITECTURE
CORPORATE SYSTEMS - Defined by Sites and Organizational Structure
HR, Procurement, Finance and Accounting, IT ManagementANALYTICS
HR, Procurement, Finance and Accounting, IT ManagementPLANNING
HR, Procurement, Finance and Accounting, IT ManagementEXECUTION
Quality, Environment, Health, Safety, Energy, Sustainability, Risk, AssetsANALYTICS
Quality, Environment, Health, Safety, Energy, Sustainability, Risk, AssetsPLANNING
Quality, Environment, Health, Safety, Energy, Sustainability, Risk, AssetsEXECUTION
MANAGEMENT SYSTEMS - Defined by Sites and Organizational Structure
ANALYTICS
PLANNING
EXECUTION
Marketing Sales
Sales
Sales
Engineering
Engineering
Engineering
Manufacturing
Manufacturing
Manufacturing
Warehousing
Warehousing
Warehousing
Distribution
Distribution
Distribution
Retail
Retail
Retail
Service
Service
Service
Suppliers AssetManagement
AssetManagement
AssetManagement
Suppliers
Suppliers
Marketing
Marketing
VALUE CHAIN SYSTEMS - Defined by Sites and Organizational Structure
Big Data Analytics, Collaboration, and Mash-Up Apps
Connectivity and Data Model
ANALYTICS & APPSANALYTICS & APPSANALYTICS & APPS
SUPPLIERS OPERATIONS CUSTOMERS & PRODUCTS
EDGE ANALYTICS AND APPLICATIONS
EDGE ANALYTICS AND APPLICATIONS
EDGE ANALYTICS AND APPLICATIONS
EDGE ANALYTICS AND APPLICATIONS
EDGE ANALYTICS AND APPLICATIONS
EDGE ANALYTICS AND APPLICATIONS
EDGE ANALYTICS AND APPLICATIONS
EDGE ANALYTICS AND APPLICATIONS
EDGE ANALYTICS AND APPLICATIONS
CELL 4CELL 3
CELL 1 CELL 2
Controller
Drive
I/O
HMI
Plant Data Center /
Application Server
SMART CONNECTED DEVICES AND ASSETS
CONNECTIVI
TY
DATA
FUNCTIONS
SETU
P AN
D CO
NFIG
URATION
SECURITY
HARDWARE SETUP& INSTALLATION
Plant Data Center / Application Server
Core Switches
DistributionSwitch
Instrumen-tation
External DMZ/FirewallExternal DMZ/Firewall
PLANT DEMILITARIZED ZONE
ENTERPRISE
Plant Data Center /
Application Server
Gateway to Plant or
IoT Network
Control
Mobile Device
ENTERPRISE
PLANT
Firewall
Firewall
Firewall
INTERNET
14
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
A New Model for Operational Architecture (Cont.)
Another important area of Operational Architecture that has not been
traditionally managed in conjunction with enterprise applications is
the Level 2 Operational Architecture for the edge analytics and con-
nectivity across devices and assets in operations.
At this level, industrial companies need to manage both the net-
working and automation infrastructure that supports security and
the flow of data with context through the traditional control system
hierarchy, as well as next generation IIoT
protocols and gateways.
This is an area where the most innova-
tion and transformation is occurring. Many
industrial companies fear that the move-
ment towards IIoT technologies involves
the movement to exclusive cloud and
gateway use. However, for the vast majori-
ty of companies it will be a controlled and
hybrid model for the foreseeable future,
where information still flows in the tra-
ditional approach but also flows through
these new flattened hierarchies. What this means is industrial com-
panies will need a data and connectivity model that harmonizes
across device, gateway, on-premise, and cloud. It also means that the
plant floor, now as much as ever, needs a cost-effective, redundant,
and fault-tolerant connectivity, compute, and storage environment
that supports the move to IIoT.
15
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
A New Model for Operational Architecture (Cont.)
At the most detailed view, or Level 3 Operational Architecture, indi-
vidual and specific elements of Level 2 will be integrated. Examples of
this could include the specific pieces of functionality that are included
within Manufacturing Operations Management (MOM) or the spe-
cific edge analytics, applications, security, device management, and
communication protocols that are used for smart connected devices.
MANUFACTURING OPERATIONS MANAGEMENTFuture: Integration and Collaboration Platforms
SMART CONNECTED DEVICESDATA
FUNCTIONS
SETU
P AN
D CO
NFIG
UR
ATION
SECURITY
Standards, Proprietary
CONFIGURATION:Platform Services, Modules/Apps
MODULES/APPS:Execution, Tracking
MODULES/APPS:Asset Tracking, MRO, RCM
MODULES/APPS:OEE, Quality
MODULES/APPS:Scheduling, Dispatching
MODULES/APPS:Time & Attendance, Training
MODULES/APPS:Purchasing, Warehouse
MODULES/APPS:EMI / OI, Reporting
ApplicationIntegration
Security& Access
Unified Asset& Production Model
Unified Operations Database & Historian
Global Deployment& Licensing
Integrated DevelopmentEnvironment
Collaboration& Workflow
Visualization & Mobility
COMMON APPLICATION FUNCTIONALITY PROVIDED BY MOM PLATFORMS:
Enterprise Applications
Industrial Automation
ESB, Standards
16
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
The Difference Between “a Lot of Data” and “Big Data” in Manufacturing
LNS Research recommends taking a relatively generic IT view of what
Big Data is and then applying the definition to the industrial space.
One definition that has received broad acceptance is the three V’s of
Big Data:
• Volume
• Velocity
• Variety
In the industrial space we have typically had to deal with large
volumes and velocity of data. According to Boeing, the 787 produces
approximately half a terabyte of data per flight. What we have not had
to deal with is variety. All of this data has been relatively well struc-
tured process data stored as time series or transactional data stored as
structured data in enterprise applications.
With the advent of the IIoT, data might include images, video, un-
structured text, spectral (such as vibration), or other forms, such as
thermographic or sound. As all of these data types come together,
industrial companies will truly have to deal with Big Data in Manufac-
turing, which will bring together a whole new set of analytics opportu-
nities and challenges.
BIG DATA
Connectivity and Data Model
17
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
The Difference Between “a Lot of Data” and “Big Data” in Manufacturing (Cont.)
As with defining Big Data, LNS Research also recommends taking
a relatively generic IT definition of analytics and applying it to the
industrial space. In the industrial space, even before Big Data,
companies were doing the full spectrum of analytics, including
descriptive, diagnostic, predictive, and prescriptive. Traditionally
these analytics have been focused on analyzing structured and
time series data to address the key drivers in industry: quality, pro-
duction, assets, delivery, innovation, and more.
Examples include:
• Descriptive: Metrics and Scorecards for Overall Equipment
Effectiveness (OEE), On Time Delivery (OTD), Scrap, Mean
Time to Failure (MTTF)
• Diagnostic: Reliability engineering, quality engineering,
root cause analysis
• Predictive and Prescriptive: Modeling and simulation,
statistical process control, advanced process control
However, as new solutions have emerged to manage Big Data,
new analytics have also emerged that are mainly targeted toward
predictive and prescriptive analytics. To add confusion, even though
these tools were developed to analyze Big Data, they can be used
with any data set: small, large, or big.
A common example of these new Big Data Analytics includes
Machine Learning, among many others. These new analytics are all
generally data focused, where the traditional tools are model and
process specific, which adds to the challenges in bridging the gap
between data scientists using Big Data Analytics and engineers using
traditional model based analytics.
Next generation Prescriptive Analytics are really about moving
beyond choosing what to do next, and optimizing operations and en-
abling innovation. In the next section we will examine the adoption
of these tools and how industrial companies can hasten the value
captured from using them.
BIG DATA ANALYTICS FRAMEWORK
DESCRIPTIVE DIAGNOSTIC PREDICTIVE PRESCRIPTIVE
What happened
What willhappen
What actionto take
Why it happened
19
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
What Are the Top IIoT Use Cases of Today and Tomorrow?
When considering adoption of IIoT and Big Data Analytics, it is infor-
mative to start with the use cases.
When it comes to the IIoT use cases being pursued today, there is
no single use case coveted by the majority of companies; instead, it
is spread quite evenly across the traditional drivers in the industrial
sector: energy, reliability, quality, production, etc. When comparing
the use cases of today vs. one year from today there are some inter-
esting insights. First, remote monitoring is top across both. Second,
energy efficiency is viewed as low hanging fruit and something more
likely to be pursued today than in a year. Finally, business transforma-
tion initiatives are viewed as a longer term use case and more likely
to be pursued in a year from now rather than today.
Remote monitoring
Energy efficiency
Asset reliability
Quality improvement
Production visibility
Internet enabled products
Business model transformation, e.g. selling capacity
Asset and material tracking
Traceability and serialization
Customer access to information
Improving safety
Supplier visibility
Improving environmental performance
0% 5% 10% 15% 20% 25% 30% 35% 0% 5% 10% 15% 20% 25% 30%
29% 26%25% 23%
24% 22%
23% 21%23% 21%
22% 20%19% 19%
19% 18%18%17%
15% 15%12% 12%
6% 8%
5% 5%
What are the top IIoT use cases your company is pursuing today?(N=252, all respondents)
What are the top IIoT use cases your company will start pursuing in the next year?(N=249, all respondents)
Remote monitoring
Asset reliability
Business model transformation, e.g. selling capacity
Asset and material tracking
Quality monitoring
Customer access to information
Production visibility
Energy efficiency
Internet enabled products
Traceability and serialization
Supplier visibility
Improving safety
Improving environmental performance
20
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
IIoT Data Sources and Types Used Today
Although many industrial companies are looking for IIoT technol-
ogy to enable some previously unconsidered use case, as the data
shows, most companies are pursuing already existing and previously
unsolved problems instead. In many cases these are quality, manu-
facturing, and reliability issues that have plagued organizations for
years. Examples of these specific use cases include:
• Dead on arrival quality issues that slip through the finished
product and functional testing
• Engineering and manufacturing tolerances that are too tight
or loose, allowing failures to the field or keeping good
products from customers
• Unscheduled downtime occurring as an unknown failure due
to systemic issues and relationships that are not immediately
obvious, like component suppliers, manufacturing processes,
environmental conditions, and customer use scenarios
It is these earlier use cases that will allow for later transformation,
like transitioning from selling assets to capacity and providing value
added connected services.
Fortunately, because of this use case progression, industrial com-
panies can start slowly with the new data sources. As is shown in the
below graph, most companies can do quite well just by collecting
Manufacturing Execution System (MES) and Programmable Logic
Controller (PLC) data to start. Then as maturity increases, they can
add in data from smart devices.
MES, quality system, or other high level software
Individual controllers (PLC)
Complex equipment with embedded control
Individual smart devices
Other
Individual “dumb” devices (sensors, switches, analogue readings)
0% 10% 20% 30% 40% 50% 60% 70%
What information from the plant are you combining for Big Data Analytics? (N=30)
67%
47%
13%
13%
13%
7%
21
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
Data Architecture Today
Surprisingly, even though most companies have not yet moved to
start broadly collecting data from non-traditional data sources,
already a third of companies are moving to a non-traditional, non-hi-
erarchical approach to data flow.
Although companies are not ripping and replacing existing control
to information system hierarchies, many companies are beginning to
deploy gateway to cloud architectures, at least in a limited capacity,
to begin delivering IIoT data to higher level enterprise applications
and analytics packages.
It remains to be seen if this information flowing through gateways
and to the cloud will be enabled on the “thing” side of gateways
with direct connections to existing automation equipment, or via a
secondary system of sensors and connectivity. There are pros and
cons to both and each will likely persist for the immediate future as
industrial companies experiment.
For just one example of the confusion, much of the data coming
from new sensor-to-gateway-to-cloud solutions is measuring data
points that are already collected within the control system but cur-
rently lack the context of the control system. But on the flip side,
these new sensor-based solutions are exclusively focused on deliver-
ing value from the new data coming from these sensors, deploy more
quickly and easily than with existing automation solutions, and often
provide a positive short-term ROI.
Today, how are you architecting the flow of IIoT data? (N=167, companies with IIoT initiatives)
Through traditional control andinformation system architecture
Primarily through traditional control and information system architecture with some use
of edge analytics, gateways, and cloud
Even split between traditional control and information system architecture and use of
edge analytics, IIoT gateways, and cloud
Primarily use of edge analytics, IIoT gateways, and cloud with some use of traditional control
and information system architecture
Other
Exclusive use of edge analytics, IIoT gateways, and cloud
0% 5% 10% 15% 20% 25% 30% 35%
34%
28%
17%
10%
8%
4%
22
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
Data Connectivity and Data Ownership
As new data sources and systems come online, emerging questions
of data ownership and data sharing are becoming critical.
One of the most important questions is “who owns the new
machine data?” The maker of the asset/device, or the user? Interest-
ingly, we see the market split on this point today, and there may not
necessarily be a verdict or even a single answer any time soon. There
are, however, a few points becoming clear.
1. When customers do not own the data, they prefer not to pay
for the raw data coming back. Rather, they want to pay for the
value services being delivered back to them that may only have
been possible through data sharing.
2. Use of the machine matters. In scenarios where the use
of the machine has no competitive differentiator, like in com-
pressed air for example, data sharing and selling compressed
air instead of compressors is not an issue. When the use of the
machine does create competitive advantage, like with CNC ma-
chines or oil field service equipment, asset users are much more
protective of data.
Who owns the data coming from the machines you deliver to customers?(N=66, machine builders only)
The customer owns the data, we are data custodians
We own the data and do not share raw data with customers
We own the data but provide raw data access to customers
0% 10% 20% 30% 40% 50% 60%
50%
29%
24%
23
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
Using Big Data Analytics
One of the most surprising results from the survey came from the
question asked regarding analytics expertise. The most common
response, representing 40% of the market, was the belief that com-
panies already had all the needed analytics expertise.
Given the data shown above regarding the IIoT use cases, new
data sources, and changing system architecture, it is unlikely this
many industrial companies actually have the right degree of analyt-
ics expertise.
We have strong analytics teams that will not require much expansion
We use or will use large scale consulting companies with specialist industry knowledge
Don’t know - This is a potential stumbling block
Don’t know - We’ll worry about this later
We plan to hire specialists in industrial analytics
We will use expert consultants fromour analytics software vendor(s)
0% 5% 10% 15% 20% 25% 30% 35% 40% 45%
From where does your company get or plan to get its analytics expertise?
40%
23%
17%
17%
13%
10%
24
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
Using Big Data Analytics (Cont.)
What is more likely, as shown by this companion question, is that
most industrial companies are just doing descriptive analytics on
structured data sets rather than predictive and prescriptive analyt-
ics with Big Data.
To help address this apparent lack of understanding, industrial
companies need to invest in both the appropriate technology, but
more importantly, process and training as well. Just as Six Sigma
and Lean were built into the fabric of continuous improvement
initiatives and packaged for subject matter experts to use the ap-
propriate financial modeling, process optimization, and variability
reduction analytics without being a statistician, Big Data tools like
Hadoop and Machine Learning need to be packaged and made ac-
cessible to industrial subject matter experts, not just data scientists.
Algorithms used in analytics system
Trend analysis
Data visualization
Statistical distribution analysis
Statistical process control (SPC)
Optimization
Regression analysis
Predictive modeling
Material performance
Correlation analysis
Simulation
Condition based monitoring
Machine learning
Data mining algorithms
Sentiment analysis
0% 10% 20% 30% 40% 50% 60% 70%
59%
44%
41%
41%
33%
30%
26%
22%
22%
19%
11%
19%
11%
7%
25
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
L5
L0
L1
L2L3
SMART CONNECTED ENTERPRISE
L4
IIoT Enabled Next-Gen Systems
MATERIALSAND
SUPPLIERS
PRODUCTSAND
CUSTOMERS
Enabling Smart Connected Assets and Smart Connected Operations
In LNS Research’s aforementioned eBook, “Smart Connected Op-
erations: Capturing the Business Value of the IoT,” we first hypothe-
sized that IIoT Platform technologies would precipitate the flattening
of the traditional hierarchical model. With this new research, we are
seeing the first quantitative evidence of this transformation occur-
ring, and as such there will be several key changes including:
• A transformation of control system hierarchy to move from one
of distributed controllers and centralized control to truly distrib-
uted control with the enablement of Smart Connected Assets.
• A transformation of MES to become an orchestra-
tion and optimization platform for Smart Con-
nected Operations, not simply an integration
and analytics middleware layer for execution
and compliance.
• A transformation of enterprise applications to more closely
map to operations instead of accounting models, and have
the ability to work flexibly with operational data and not just
structured transactional data.
• The enablement of mash-up applications and analytics that
can enable Big Data from anywhere to anywhere and support
true end-to-end value chain processes.
27
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
IIoT Challenges
There is not an industrial company in business today that would not
like easier access to operational data and the decision support tools
to better address quality, production, and reliability issues.
This is why for many it is counterintuitive that the top two chal-
lenges in IIoT adoption are not technical, but rather funding and
business case development
The reason is a classic catch 22: before Big Data Analytics are
implemented, companies cannot accurately predict their benefits.
Likewise, without a comprehensive understanding of Big Data ben-
efits, companies are reluctant to invest the time and resources on
their implementation.
Funding
Building a business case
Understanding what IIoT is and how it applies to your business
Security
Standards
Finding the right technology partner(s)
Gaining insight from Big Data
Developing new IIoT software applications
Company culture
Data gathering from legacy systems
Product design and development complexity
Hiring the right talent
Executive support
Scaling to 1,000s or 1,000,000s of devices
0% 5% 10% 15% 20% 25% 30% 35%
32%30%
26%25%
22%17%
16%14%14%
13%12%
8%8%
5%
What are the top challenges your company faces in deploying IIoT technology?(N=269, all respondents)
28
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
A Business Case Journey that Aligns to Strategic Objectives and Maturity
To address this catch-22 companies should think of the Big Data
Analytics investment as a journey that is based on Operational Ex-
cellence maturity, scope, and metrics rather than a one-off ROI cal-
culation. Operational Excellence maturity is the driving factor for
increasing the scope and value of the business case. LNS Research
recommends using a 5-Level approach to quantifying maturity,
where at the lowest, Ad Hoc Level companies are unable to meet
the current and future demands of customers and at the highest,
Market Leader level companies are able to define and transform
markets, disrupting incumbents.
The following matrix will allow companies to evaluate their
current position based on their capabilities.
Globally integrated and Harmonized. Fully embracing
emerging capabilities
Predictive, role-based,real-time metrics connected
to corporate goals
Predictive, role-based,real-time metrics connected
to corporate goals
AD HO INNOVATION LEADERC CON. PRO. AGI.
STRATEGY& EXECUTION
LEADERSHIP& CULTURE
ORGANIZATIONALCAPABILITIES
BUSINESS PROCESSEXCELLENCE
TECHNOLOGYCAPABILITIES
PERFORMANCEMANAGEMENT & KPIs
Disconnected and disparate
Disconnected and disparate
Non-role based, manual KPIs, disconnected
from corporate goals
Disconnected from corporate objectives
Fully integrated with corporate objectives
Operational Excellence is a department rather than shared responsibility
Operational Excellence fully integrated into corporate structure
Operational Excellence is integral part of leadership
and culture.
Operational Excellence distinct from corporate structure. Not in goals or incentives
INNOVATION LEADERDrives standards and expectations
AGILEEvolved people, process, and technology across the enterprise
HARMONIZEDFlexibly unified at the organizational level
CONTROLLEDRepeatable within organizational, process, and/or technology boundaries
L1L1
L2L2
L3L3
L4L4
L5L5
AD HOCUnstandardized with significant variation
29
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
A Business Case Journey that Aligns to Strategic Objectives and Maturity (Cont.)
When the business case is viewed as a journey, most industrial
companies should begin on the cost side of the equation and within
specific functional areas of operations. With limited maturity, there
is not a shared vision of how productivity gains will drive actual
financial benefits. By starting with a single area like quality, manu-
facturing efficiency, asset reliability, or energy usage, the need for
collaboration is minimized and cost reductions clearly go to the
bottom line, eliminating uncertainty of real results.
As maturity increases and initial cost reduction benefits are re-
alized, the scope of the business case can increase and the types
of metrics measured can move to being value based. As more and
more maturity is realized, industrial companies can more accurately
predict the economic benefits that will be realized from productivi-
ty gains and, ultimately, the achievement of strategic objectives like
business model transformation or the entry to new markets.
5$$$$$
VALUE CENTERCOST CENTER
Siloed
BUSINESS CASE AND OBJECTIVE SCOPE
OPER
ATIO
NAL
EXCE
LLEN
CE M
ATUR
ITY
METRICS
BUSINESS CASEJOURNEY
DEPARTMENT EXECUTIVECROSS-FUNCTION
Operational
Financial
Value**
Big Data**Big Data Analytics, Diagnostic, Predictive, Prescriptive
**e.g. Revenue and Earnings
1
2
3
4
$$$$$
$$$$$
$$$$$
$$$$$
30
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
Recommended Actions
IMPLEMENT A BUSINESS CASE JOURNEY FOR BIG DATA ANALYTICS
Map your organization’s business case journey for Big Data
Analytics to the current and anticipated maturity of your
Operational Excellence capabilities. At lower level of maturity
you should focus on a narrow scope and cost focused benefits.
At higher levels of maturity, we recommend you focus on a
broader scope and value based metrics with direct connection
to strategic objectives (but not necessarily short-term
financial gains). Map the journey to align with an Operational
Architecture that accounts for Big Data Analytics across
operations. To improve Operational Excellence maturity,
ensure you invest in systems and training to make Big Data
Analytics accessible to existing subject matter experts, not just
data scientists.
CHOOSE AN INITIAL USE CASE FOR BIG DATA ANALYTICS THAT
ALIGNS TO YOUR COMPANY’S PAIN POINTS AND/OR COMPETITIVE
DIFFERENTIATION Often these initial cases are for quality,
manufacturing efficiency, or reliability. Quality is a great
starting point because improved quality can drive both short-
term ROI through reduction in scrap and rework, but also
long-term benefits for a differentiated customer experience
and improved product design based on quality information
coming from connected products.
IIoT Platform technologies are currently driving the most transformative
period in the industrial sector over the past 40 years. As industrial
executives attempt to establish high level strategic objectives, it is
critical that a formalized and structured approach is taken to Digital
Transformation that establishes an expanded view Operational
Architecture and captures the value of Big Data Analytics.
ESTABLISH A DIGITAL TRANSFORMATION FRAMEWORK
Establish a Digital Transformation leader and new group
responsible for a framework that connects and enables for
change all levels and functions of the organization. Incorporate
feedback loops at each stage of the journey and ensure that
high level strategic objectives are aligned with Operational
Excellence initiatives, system architectures, business cases,
and solution selection.
ESTABLISH AN OPERATIONAL ARCHITECTURE
Without a formal Operational Architecture, your organization
will not be able manage changing architectures based on new
IIoT Platform technologies and capture the potential value of Big
Data. Ensure that your organization’s Operational Architecture
includes a robust and flexible data and physical infrastructure
model that can:
o Tie together structured, semi-structured, and
unstructured data
o Manage IT and OT convergence
o Support traditional descriptive and diagnostic analytics like
dashboards, trend analysis, regression analysis, and more
o Support next generation predictive and prescriptive
analytics like machine learning
31
IIOT
AND
BIG
DATA
AN
ALYT
ICS:
Ho
w M
anuf
actu
ring
Syst
em
Arch
itect
ure
Is B
eing
Tra
nsfo
rmed
SECTION
TABLE OF CONTENTS
PAGE
1 2 3 4
lnsresearch.com
Presented by:
Connect:
IIoT AND BIG DATA ANALYTICS:How Manufacturing System Architecture Is Being Transformed
Author:
Matthew Littlefield,
President and Principal Analyst
© 2016 LNS Research.
www.rockwellautomation.com