Whitepaper

Industrial Internet of Things:

potential for the optimisation of existing plants
 Industrial Internet of Things 3

OVERALL COORDINATION: Mirko Lampe

LAYOUT: Fanni Robel

With content support from:

Whether oil and gas, chemicals, pharmaceuticals,

mining or food, the development of an Industrial
Internet of Things (IIoT) in combination with algorith-
ms and artificial intelligence, releases an enormous
optimisation potential - also and especially for exis-
ting plants. New business models transform capital
expenditures (CAPEX) into operating expenditures
(OPEX) and reduce the financial risk.
For this, however, the information chain consisting
of individual IIoT modules must not only be stringent,
but also secure, scalable and ideally capable of being
retrofitted during operation – from sensors and com-
munication infrastructure to cloud analysis and user
apps. There are also many success factors to con-
sider at an organisational level.

THIS PAPER IS INTENDED AS A GUIDE.

 Industrial Internet of Things 5

Table of contents

Introduction 6

1. Transformation: How the IIoT changes the automation market 8

2. Potential for benefits: What’s to be gained by the process industry? 12

3. The challenges so far and overcoming them 14

4. New architecture models: Freedom, flexibility and security 15

5. General solution approach for IIoT retrofitting 18

5.1. Sensor level – beacons, MEMS, actuators, … 18

5.2. Communication – networks and protocols 21

5.3. Edge processing – data pre-processing at the “edge” 26

5.4. Gateways – data collection and forwarding 27

5.5. Cloud level – local or global? 28

5.6. Analysis – big data, algorithms and AI 29

5.7. Application level – user devices and apps 36

6. Specific implementation examples 37

6.1. Beacon/ RHT sensor for the simplification of O&M 37

6.2. Track & trace of mobile assets under Ex conditions 38

6.3. Valve position monitoring 40

6.4. Corrosion monitoring of pipelines 41

7. Barriers and success factors for IIoT retrofitting in hazardous areas 42

Checklist 44

8. Summary and conclusion 45

6 IIoT GUIDANCE Whitepaper 9/2020

INTRODUCTION

T
he Industrial Internet of Things (IIoT) is probably
the most powerful tool within the Industry
4.0 movement. As a further development of
the previous M2M (machine-to-machine)
communication, the individual modules (IoT building
blocks) form an industrial Internet that can be used very
flexibly, scaled easily and secured by optional encryption1.
IIoT-GUIDANCE opens up the associated potential benefits
in the hazardous areas of the process industry based
on decades of experience and a holistic approach to
solutions including Ex-certified sensors, gateways and
other IoT building blocks, which are also suitable for
retrofitting existing plants. This makes it possible to
obtain important information that previously could not
be collected, or only at an exorbitant cost, in order to
make production processes economically transparent,
even retrospectively, and to identify competition-relevant
relationships in combination with tailored algorithms,
cloud computing and artificial intelligence2. The deepened
or broadened understanding of processes can in turn be
used to increase overall equipment effectiveness (OEE),
further enhance security or align production with market
requirements to an ever greater extent. To this end, the IIoT
also permits completely new business and service models
which gives prospects to reduce financial risk and fixed
costs for producers and suppliers.

______________________
1
See Chapter 4
2
See Chapter 5.6
8 IIoT GUIDANCE Whitepaper 9/2020

Transformation: How the IIoT changes

1 the automation market
As “disruptive” technology, the IIoT is well on its way to questioning the previous busi-
ness models and changing the automation market. More than seven billion IoT devices
are already active today. By 2025, this figure will rise to 22 billion3, with the number
of devices for the Industrial IoT being many times higher than the consumer market4.
Market analyst Bain & Company expects total sales of devices (things), network tech-
nology and gateways, cloud and analytics as well as apps and services5 to double to
520 billion US dollars6, and IoT Analytics experts even expect 37 percent growth to 1.6

Global IoT Market Forecast

1.567
1.600
GLOBAL IOT MARKET IN $B

1.400

1.200
%
39
GR
1.000 CA

800

600

400
151
110
200

Sales development
of the overall IoT market 2017 2018 2019 2020 2021 2022 2023 2024 2025
(Source: IoT Analytics).

The Internet of Things is not a single market

apps and services

cloud and analytics CONSUMER ANALYTICS

network, directory NETWORK AND GATEWAY

and gateway

ENTERPRISE
things AUTONOMOUS AND INDUSTRIAL

Not a single market: the IoT

market consists of several
overlapping eco-systems. consumer enterprise and industrial
(Source: Bain & Company)

______________________
3
State of the IoT 2018: Number of IoT devices now at 7B – Market accelerating. IoT Analytics 2018
4
The Industrial Internet of Things, PwC 2016
5
https://www.bain.com/insights/defining-the-battlegrounds-of-the-internet-of-things/
6
https://www.bain.com/insights/unlocking-opportunities-in-the-internet-of-things/
7
https://iot-analytics.com/state-of-the-iot-update-q1-q2-2018-number-of-iot-devices-now-7b/
 Industrial Internet of Things 9

Not only the large cloud and service providers participate in this, but also players
in the analytic space such as Trendminer and finally the hardware manufacturers.
By gradually supplementing product sales with remote support and other soft-
ware-based value-added services (software “as a Service”, SaaS), they can ultimately
act as platform providers (PaaS) and offer their customers device management, pre-
dictive maintenance or IoT analyses as a platform or service. Holistic service providers
are taking on a new role in the course of this development: they are becoming media-
tors and moderators between the business areas of automation, IT and management.
IIoT-GUIDANCE serves the industry to enable the best possible results and a fast ROI.

Building platforms product-based offerings to service-based offerings by building platforms.

platform as-a-service

product + value-add service

product + remote support

CUSTOMER ENGAGEMENT
REVENUE EXPANSION

stand-alone product

IOT SERVICES ROADMAP

New business models: increase in sales and customer loyalty through the development of IoT services and platforms.
(Source: PwC)

New business and service models

In addition to the established cloud and service providers, hardware manufacturers
are also participating in the new sales potential. By supplementing their products
with remote support and other software-based value-added services (software “as a
service”, SaaS), they are gradually transforming themselves into platform providers
(PaaS) that offer their customers device management, predictive maintenance or IIoT
analyses, for example. Companies like IIoT-GUIDANCE are taking on a nessesary new
role: they are becoming mediators and moderators for their customers between the
business areas of automation, IT and management. At the same time, IIoT-GUIDANCE
with the network of partners functions as an IIoT enabler also for hazardous areas.
The user centric integrated and holistic approach enables step-by-step implementa-
tions with best possible results and fast ROI. Please also see graphic at page 28.
10 IIoT GUIDANCE Whitepaper 9/2020

The most important IIoT services for the process industry:

With their new services, IIoT providers help the process industry to meet a wide range of
challenges. Here are the most important business models:

1. Process optimisation 4. Asset tracking

Processes, plants, machines and logis- In combination with positioning ser-
tics are made transparent by the IIoT. By vices, vehicles can be located and the
analysing the collected data, physical transport of materials, products and
security, data security and efficiency equipment monitored and traced.
can be improved, particularly in terms of In this way, errors or damage can be
time, energy consumption and the use tracked, occupational safety or logistics
of materials. improved, and shutdown procedures
optimised.

2. Predictive maintenance 5. Automatic fulfilment

Condition monitoring of plants and Analogous to the „Amazon Dash
machines and subsequent analyses Button“, stock levels (material, spare
with artificial intelligence make it parts, etc.) can also be monitored in a
possible to detect immanent failures professional environment through iden-
in good time. Maintenance can thus be tification and localisation and reordered
performed with foresight, instead of at automatically.
fixed intervals, as has been the case so
far. This increases the serviceable life
and reduces costs.

3. Remote diagnostics & support 6. Compliance monitoring

Through Internet-based remote main- Last but not least, the IIoT can also be
tenance, experts solve problems quickly used to comply with regulations and
and competently from a distance, either guidelines. This applies on the one hand
through direct access via a secure to legal regulations on environmental
connection, or by supporting local staff. protection and occupational safety,
Using data headsets or tablets, critical but also to specific safety regulations
situations can be better assessed (e.g. such as access monitoring in hazardous
temperature, humidity, condition of the areas.
machine, plant and process).
 Industrial Internet of Things 11

predict
reduce costs,
maintenance
while improving
based on process
product quality
data

see if similar predict

behavior happened heat exchanger
before fouling

monitor find root causes

process behavior through pattern
real time recognition

solve 80% of your

send early
day-to-day
warnings to the
production
contol room
questions

determine
optimize
best operating
energy
zones based
efficiency
on process data

Examples of process optimisation through IIoT Analytics. (Source: Trendminer)

New financing models

Another advantage of IIoT development for producers in the process industry is that
they can use the new service models to convert CAPEX into OPEX and at the same time
minimise the financial risk. Because the production performance of machines and
systems becomes transparent, the manufacturer or operator can invoice them on a
usage basis, for example for motors or pumps by the minute. This pay-per-use model
represents a new alternative to traditional financing models. Operators thus have
comparatively little or no investment to bear. The financial
risk is minimised accordingly. In the future, it is even con-
The IIoT extends
ceivable that suppliers, vendors or contractors will provide
hardware and software free of charge, but will share in the value chain from
the resulting savings. IIoT projects could thus become hardware to value-
a win-win situation for suppliers and operators, with all
parties achieving the best possible results.
added services. For
the operator, CAPEX
becomes OPEX.
12 IIoT GUIDANCE Whitepaper 9/2020

Potential for benefits: What’s to be gained by

2
the process industry?
According to a study by the ARC Advisory Group, unplanned downtime costs the global
process industry 20 billion US dollars annually8, 80% of which is due to process-re-
lated operational disruptions. Until now, there has been a lack of efficient technolo-
gies to detect them. The IIoT now provides the means to gain insight into the existing
processes, also retrospectively. With the help of modular IIoT building blocks, new and
existing plants can be made transparent with an acceptable level of investment in
order to increase overall equipment effectiveness (OEE) and release hidden capacity
reserves in all three OEE disciplines: availability, performance and quality9.

Overview: Potential benefits for the process industry

Service capacity Explaination/Examples

Increase plant
e.g. through condition monitoring and analyses for predictive maintenance
availability

Increase degree Continuous process improvements through data capture, analyse and resultant
of performance optimisations

Increase degree Constant / reproducible product quality, e.g. through monitoring quality-critical data and
of quality early notification

Increase
Optimum exploitation of the production plants and logistics (supply and demand chain)
productivity

New planning and control possibilities for production by combining process, ERP and
Increase
environmental data through to production in batch size 1 (modular type package, “one of
flexibility
production”)

Increase process Improved understanding of plant dynamics as well as of material and product quality, e.g.
understanding forecast of the crude oil grade

Market-dependent Planning depending on the market development. For example, to determine the
production planning production volume of oil and gas as dependent on product quality and market price

Improve learning
Self-learning systems and AI for continuous improvement
capacity of a machine

Optimise investment Transform investment expenditure (CAPEX) into operating costs with the assistance of
deployment service providers (OPEX)

Increase plant
For example, by monitoring and reporting deviations from standard values
and human safety

Balance out skilled employee shortage, for example through sound remote assistance,
Know-how transfer assistance systems, troubleshooting capability, transferring employee expertise to digital
processes

Efficient entry Fast entry via pilot project with ROI < 1 year, foundation for further iterative digitisation
to digitisation steps

______________________
8
https://www.industry-of-things.de/schluss-mit-stillstandszeiten-a-650427/
9
https://www.business-wissen.de/artikel/oee-auf-der-suche-nach-den-verborgenen-kapazitaetsreserven/
 Industrial Internet of Things 13

New approach to the subject of digitalization

Thanks to its high flexibility and scalability, the IIoT allows a new, agile approach
to digitisation in hazardous areas. This is because the ATEX-compliant modules
including sensors, network technology, gateways and edge computing can be easily
retrofitted during operation and allow additional data to be collected and evaluated
simply even in relatively complex and rigid systems. Initial
pilot projects are designed in such a way that the ROI is
achieved in less than a year. The experience gained serves Digitalization is
as a basis for further projects. Small iterative steps thus a journey; the IIoT
take the place of the „very large digitisation solution“, which
is financially, organisationally and economically unrealistic.
is the vehicle.
Instead of working towards this for years, the idea is to first
create transparency for a selected task with the help of a service provider and then to
develop further opportunities, for example for a larger retrofit or other smaller sub-
tasks. In this way, digitalization expertise grows continuously, while the time required
and the ROI duration of the next projects decrease (see checklist in the appendix).
14 IIoT GUIDANCE Whitepaper 9/2020

The challenges so far

3
and overcoming them
Why could this attractive potential not be exploited earlier? This is mainly due to the
lack of the necessary technology building blocks, including computing and storage
capacities, software and efficient possibilities for wireless and wired data transmis-
sion10 as well as powerful and yet energy-efficient mobile hardware with high battery
density for long wireless operation11. Accordingly, automation focused primarily on
process control and safety. Because irrelevant plant components are
The problem: the installed not wired or do not support external data transmission, the control
loop data cannot be further communicated, analysed and linked. As a
sensors are not IIoT-capable; result, the majority of the sensors currently used is not IIoT-capable
data cannot leave the control and therefore cannot be used to optimise processes and occupational
health and safety. Subsequent changes, such as cabling, are either
loops. Control units may not very cost-intensive or not permitted at all, for example in the case of
be altered. Only certified process or safety-critical controls or certified pharmaceutical plants.
hardware is permitted in In addition, data collected in control loops cannot be further commu-
nicated, analysed or linked with other data. Last but not least, only
hazardous areas. certified hardware is permitted in hazardous areas. What is more, if a
fire certificate is necessary for their operation, the hardware may not
be installed in ongoing operations, but only during the next plant shutdown, costing
valuable time. Another fundamentally important issue is that of IT security, reliability
and latency, which must of course also be met by wireless solutions12.

USA / NORTH AMERICA EUROPE RUSSIA / CIS CHINA

Standards: NEC Standards: EN Standards Standards: IEC/EN, Local
Certificates: UL, FM, CSA Certificates: PTB, TÜV... Certificates: Custom Union Certificates: Local (Basis ATEX,
Russia, Belarus, Kazakhstan IECEx)

ATEX

IECEx

NEC

IECEx / ATEX / Local

IECEx / NEC / Local

CSA

BRAZIL / SOUTH AMERICA AFRICA MIDDLE EAST / INDIA AUSTRALIA / ASIA PACIFIC
Standards: IEC, NEC, Local Standards: EN, IEC, NEC Standards: EN, IEC Standards: IEC
Certificates: IINMETRO (Basis Certificates: mixed Certificates: ATEX, IECEx, PESO Certificates: IECEx
IECEx, UL...)

Use and acceptance of international Ex approvals. (Source: Jan-Rieks Zonderman (iSafe mobile)

______________________
10
See Chapter 5.2.
11
See Chapter 5.1
12
See Chapter 4
 Industrial Internet of Things 15

New architecture models: Freedom,

4
flexibility and security
Thanks to the new possibility for energy-efficient collection and secure processing
of data, the IIoT is an ideal complement to classic automation. Three architectural
approaches offer the process industry a high degree of freedom and flexibility for
implementation:

1. Developing new automation with already integrated IIoT functionality

(greenfield)
2. In addition: existing automation and retrofitted IIoT. PLC systems and IIoT In the brownfield, the IIoT works in
feed a common database13 (brownfield) addition to or parallel with production
and enterprise IT.
3. Parallel: IIoT works as an independent system with separate data collection *Source: https://www.industr.com/de/
zuverlaessiger-datenfluss-in-der-
and processing (brownfield) digitalen-prozessanlage-2331922 /
**Source: EMERSON AG

 Storage *  Algorithm development

IIoT Cloud  Aggregation  Integration in existing system
at customer / operator
 Processing
 Logic
 Presentation
 Independent from process control / system approach
 Management
?

Traditional Pathway wired wireless IIoT Application Pathway

Hand made controlling Using Enterprise Mobility Fully connected plant

Still reality Reality & short term Mid term & Future

Communication flow bi – directional Communication flow one way**

High precision sensors / real time Simple sensors / information if changes & timestamp

Expert
Simple
Data Field
Sensors
Sensors

PROCESS / AUTOMATION MAINTENANCE

*Source: https://www.industr.com/de/zuverlaessiger-datenfluss-in-der-digitalen-prozessanlage-2331922 /
**Source: EMERSON AG

AUT solution IIoT solution

Sensor Corded / i. e. 4... 20mA Corded / Wireless / battery powered

Precision and response time high typically lower*

Connectivity / Protocol

Gateway Corded Corded / Battery powered

Data Flow bi-directional one direction / read only**

ERP or cloud update rate realtime / nearly realtime some minutes or threshold based***

Sales Modell Traditional (product) SaaS / PaaS / Leasing / Pay per use

*
compared to corded in process industry
**
bi-directional only for updates / security patches over the air / depens on the IIoT solution
***
only if changes / timestamp

______________________
13
Data diodes isolate the IIoT from automation, see Namur Open Architecture (NOA)
16 IIoT GUIDANCE Whitepaper 9/2020

IIoT retrofit without shutdown

One of the revolutionary features of the IIoT is therefore that it can also unfold its
potential benefits independently of automation and can be installed without a shut-
down. An analogy from the automotive area illustrates this: here, older vehicles can
be retrofitted with a tyre monitoring system regardless of the extent of the existing
sensors. To this end, sensors are simply screwed on to the tyre valves, which measure
the tyre pressure and transmit this data to a simple control unit where the information
is processed and visualised.
BROWNFIELD

real time monitoring

beeper

wireless transmission

2 min DIY installation

Simplification!

VEHICLE SENSORS ≈ 100 SENSORS TODAY

GREENFIELD

Vehicle analogy: modern car with

numerous sensors (greenfield) and lane departure system | front object CCD camera | front airbag sensors | ASCD | nightime pedestrian warning |
by comparison the retrofitting for drowsiness sensors | front object laser radar | nightime pedestrian IR sensor | active park assist | tire pressure
sensor | rear object monitor CCD camera | rear camera | side curtain sensor | blind spot detection | cross traffic
older vehicles (brownfield): simple
alert | cental computer | rear object laser radar | wheel speed sensor | tire pressure sensor | collision sensor | side
retrofit capability for a wireless
airbag SRS | adaptive cruise control | steering angle sensor | automatic brake activator | etc.
tyre pressure system
(Source: IIoT-GUIDANCE)

Retrofitting data functions securely

Similarly, plants in the process industry can be retrofitted with wireless, Ex-certified
sensors in order to be able to use new IIoT functions. The so-called NAMUR Open
Architecture (NOA)14 was developed in order to achieve the systemic openness
required for this without endangering the elementary requirements of availability,
function and safety. It describes, among other things, how the data of intelligent
sensors and actuators are to be read and evaluated from the field level to the cloud.
So-called “data diodes” as separating instances between automation and IIoT ensure
that they flow exclusively “from inside to outside” and that the automation systems are
not influenced in any way.
______________________
14
https://www.namur.net/de/fokusthemen/namur-open-architecture.html
 Industrial Internet of Things 17

Companies moving from product-based offerings to service-based

offerings by building platforms.

CENTRAL M+O

historian central HMI

further 4.0
use case
production advanced
scheduling
network simulation analytics

verification
of request
PLANT SPECIFIC M+O
CORE PROCESS CONTOL

OPC UA engineering
advanced process alarm
control manegement
highly reliable
HMI IT infrastructure
4.0 device
management dispatching
deterministic
control system
DCS/PLS
app platform

data direction
control
low-cost vibration TC 4711
multi sensor 4.0 out FC 4713
open interfaces
PRODUCTION PLANT 4-20mA / remote IO / fieldbus / wireless / ethernet in the field
proprietary
interfaces

Industry 4.0 enabler: the standard-based, adaptive NAMUR Open Architecture (NOA) is simple to integrate. Data
diodes (Data Direction Control) separate the area of monitoring and optimisation from the core applications
(Source: PROFIBUS Nutzerorganisation e.V.)

Data diodes separate

the IIoT from the
automation and ensure
smooth and safe
operations.
18 IIoT GUIDANCE Whitepaper 9/2020

General solution approach

5
to IIoT retrofitting
What does a resilient solution approach to IIoT therefore look like? It must extend strin-
gently across all levels – from data collection (sensors), pre-processing (edge comput-
ing) and forwarding (gateway) to analysis (cloud) and the integration in business apps.
At the same time, the IIoT retrofit must also be easily scalable, as economical and
future-proof as possible – either as a supplementary data source for existing auto-
mation or as a completely separate parallel system. And of course, all hardware
Solution levels for components must be certified for the respective hazardous areas. In addition, other
modular IIoT retrofitting
of existing plants compliance regulations must also be observed, not least IT and data security issues
(Source: IIoT-GUIDANCE) or works council topics15.

O&G CHEMICAL PHARMA MINING FOOD & BEVERAGE

BROWNFIELD FOCUS

BUSINESS PROCESSES
NB-IoT wired/wireless

Information via:
Edge processing
depends on use Global cloud Data intelligence Mobile
Wireless vs. battery cases
Sensors depends Data concentration, battery/powerline Local cloud End user, HMI
on use cases Consultant
Pre-analysis, Gateway depends SCADA
on use cases IaaS SaaS
PaaS ERP/ MES
Integration
EX on demand EX on demand EX on demand

MODULAR SYSTEM APPROACH

Cloud
Sensors Local Edge Cloud User
Gateway Algorithms
Actuators Network Processing Operation Device App
Analysis

5.1. Sensors – beacons, MEMS, actuators, …

Installing an IIoT does not necessarily mean the use of radio technology. For example,
there are also wired pressure transmitters and temperature probes for use in hazardous
areas (classic automation) or devices with direct device to cloud communication (D2C).
However, simple, wireless sensors greatly expand the spec-
trum of retrofit solutions. Most of them can be installed
IIoT retrofits must be simple, during operation, i.e. without a time-consuming shutdown
scalable, economical and future- procedure. Wireless sensor solutions such as beacons,
proof but at the same time also Ex- which send data to a gateway at a defined interval, or ener-
gy-efficient MEMS sensors17, which measure and transmit
safe, data-safe and reaction-free. several values simultaneously, can work for years without
battery replacement using the advertise mode in the Blue-
tooth Low Energy (Bluetooth LE) radio specification18. This dramatically reduces invest-
ment and maintenance costs compared to wiring. Examples are zone 1 certified sensor
solutions for monitoring temperature, air and valve position. The gateway required for data
collection and forwarding is available for zones 1 and 2 and can therefore be selected
according to the hazardous area requirements.

______________________
15
See Chapter 6
17
Micro-electro-mechanical system. MEMS sensors concentrate several measurement functions in a very small space and work, for example,
in vehicles, fitness trackers, smartphones and virtual-reality headsets. In combination with a wireless and energy-efficient data transmis-
sion, they are also suitable for the Industrial Internet of Things.
18
See https://de.wikipedia.org/wiki/Bluetooth_Low_Energy
 Industrial Internet of Things 19

O&G CHEMICAL PHARMA MINING FOOD & BEVERAGE

BROWNFIELD FOCUS

5 min
Own cloud

BUSINESS PROCESSES
platform
RHT sensor Under
7 security development
and Accesss with customers
levels Information via:
Programming Workshop
treshold Hosting & Guidance Mobile
Valve sensor in EU
Around HMI
Customer 6 month
specific learning SCADA
version period
possible ERP/ MES
Integration

MODULAR SYSTEM APPROACH

Cloud
Sensors Local Edge Cloud User
Gateway Algorithms
Actuators Network Processing Operation Device App
Analysis

Performance and energy consumption Example of modular IIoT retrofitting

of existing plants with sensors in
Zone 1 and gateway optionally Zone 1
Sensors for 3D and infrared image recognition, which, as a self-learning system with or Zone 2 as well as connection
the aid of AI, recognise hotspots by comparison with reference images, are correspon- to cloud platform.
(Source: IIoT-GUIDANCE)
dingly more complex. IIoT-GUIDANCE network of specialists can provide solutions for
the use in hazardous areas. In order to minimise battery life, the computing work can
also be outsourced to a wired gateway (see 5.4, Edge Processing). Energy-efficient
wireless data transmission, for example via Bluetooth LE (Low Energy), as well as
sophisticated cycles of waking up, measuring, calculating, sending and falling asleep
are important for long, maintenance-free operation. In this manner, the solution
provider SAVVY® Telematic Systems AG thus achieves a battery life of up to 15 years
even in hazardous areas19. A further success factor: simple, Ex-compatible instal-
lation on the object to be monitored during operation by screwing, riveting, gluing or

Wired sensors for hazardous

areas e.g.: PT100 resistance
thermometer.
(Source: IIoT-GUIDANCE)

Wireless sensors can be

simply retrofitted during
operation.
(Images from Trisens.no)

______________________
19
Source: SAVVY® Telematic Systems AG
20 IIoT GUIDANCE Whitepaper 9/2020

What can be monitored?

The sensors form the smallest link in the IIoT information chain, with which the com-
plete system stands and falls. Therefore, it must match the respective application
precisely. Classic sensor technology is wired. As described above, wireless solutions
are particularly simple and much more cost-efficient to retrofit, which is why they are
the focus of particular attention in the following.

Whether wireless or wired - sensor technology can be used to monitor a wide range of
process parameters in products, plants and supply chains. In addition to direct meas-
urements (temperature, air pressure, humidity, light, etc.), it is also possible to derive
information. For example, simple switches or optical sensors can detect certain lever
positions. Strain sensors clamped to a carrier system of tank wagons, containers or
pipes allow conclusions to be drawn about the filling level (weight), temperature and
pressure. Position sensors provide acceleration values for vibration analysis, while
magnetic field measurements provide information about the current flow. Here is an
overview of various measured values and their application:

Overview: Potential benefits for the process industry

Measured value Application examples

Process monitoring, alarm on exceeding or falling below a threshold value, monitoring of

Temperature
overland lines

Air pressure Process and environment monitoring (threshold)

Air humidity Process and environment monitoring (threshold)

Light Automatic switching on and off, energy saving

Magnetism Monitoring of switch positions or power consumption (conclusions as to current flow)

Acceleration Plant monitoring through vibration analysis. Recognition of mechanical wear for
(inclination) predictive maintenance

Position Position of switches and levers

Asset tracking, safety monitoring and access control for certain areas or Ex zone 1 / 2;
Location (GPS)
person localisation

Battery level of the

Service planning
sensor

Identification Identification, localisation of products, ingredients and equipment

Distance Geofencing, monitoring of Ex zones

Monitoring of filling level (weight), temperature (longitudinal expansion) and pressure of

Expansion
tanks, pipes, tank wagons, etc.
Fault/damage monitoring and motion detection through intelligent comparison with
3D and infrared
reference images: Infrared: detection of hotspots
images
3D: position, clearance, speed, etc.

On/off switch Recording of valve positions (switch mechanically measures lever or switch)
 Industrial Internet of Things 21

5.2. Communication – networks and protocols

The process industry also places specific requirements on the forwarding of data,
which can be done classically and with near real-time capability via cable, for example
via fieldbus or Ethernet protocols, or wirelessly if real-time data is not required. Their
strength lies in the simple and economical retrofitting during operation. Radio com-
munication is also suitable due to the sometimes harsh environmental conditions, for
example in order to penetrate built-up areas and reach high plant sections without
wiring and to cover long distances of up to several kilometres. Accordingly, the selec-
tion of radio technology is typically based on range, data rate (amount of data per
time unit) and reliability (redundancy). The requirements for latency (50 ms to a few
seconds) and error rate (less than 10 to 5 %) are rather moderate due to the relatively
low processing speed. For efficient and secure data transmission, there are now
several radio technologies with different characteristics, including standardised
(e.g. MIOTY) and proprietary (e.g. ZigBee).

LPWAN technologies work either in the royalty-free sub-gigahertz range (e.g. at 915
MHz or 868 MHz), or in the royalty-payable mobile radio network (3G, LTE up to 5G).
A distinction can be made between the following five groups according to the
technology used20.

1. Mobile radio / professional mobile radio (e.g. Tetra/DMR)

2. Low Power WAN technologies (e.g. MIOTY, LoRaWAN / IEEE 802.11ah,
SIGFOX, NB-IoT, etc.)
3. Wireless LAN based on IEEE 802,11
4. Radio networking based on IEEE 802.15.1 (Bluetooth, WISA)
5. Radio networking based on IEEE 802.15.4 (e.g. Zigbee, ISA100.11a,
WirelessHART)

Cellular

5G

4G

WiFi 3G

2G Long range
DATA RATE

Short range VSAT

Bluetooth

BLE

6LowPan
ZigBee
NFC/RFID LPWAN
EnOcean

RANGE A comparative study of

LPWAN technologies for
large-scale IoT deployment.
(Source: www.sciencedirect.com)

______________________
20
www.industrialradio.de/Attachments/Funktechnologien_Industrie_4.0_Web.pdf
22 IIoT GUIDANCE Whitepaper 9/2020

Low-power WAN technologies

Low Power Wide Area Networks (LPWAN) are particularly suitable for typical IIoT tasks
such as telemetry or condition monitoring over medium or long distances. Since they
can transport small amounts of data with low energy consumption even over long
distances or through densely built-up areas, they close the gap between mobile radio
(3G, 4G, 5G) and near-field radio (Bluetooth, WLAN, Zigbee, etc.). LPWAN can be used
to set up national, international and geographical networks for regionally limited
tasks. A distinction must be made between cellular LPWANs that operate in licensed
mobile radio and those that use the licence-free sub-GHz range.

Sub-GHz LPWANs use protocols such as MIOTY, LoRa or Sigfox, and can either be
booked by a telecommunications operator (e.g. Sigfox) or operated by the company
itself with its own base stations or gateways. In this way, the network data can be
limited to the physical installation without connection to a service provider. However,
the data rate is lower than with cellular LPWANs that use protocols such as LTE Cat
M1 (eMTC) or LTE Cat NB1 (NB-IoT).

Software-defined standard protocols (e.g. MIOTY, but also NB-IoT), which work with
standard hardware components are particularly interesting from an economic point
of view. In the case of MIOTY or Lora, for example, an ordinary industrial PC with radio
receiver can serve as a base station. A transmitter with a sub-GHz transceiver chip
is sufficient as transmitter. The NB-IoT standard, which is also software-defined, is
based on the LTE mobile radio standard and can also be integrated into the existing
LTE infrastructure via a software upgrade. By 2025, five billion IoT mobile radio
modules based on 4G, LTE Cat-M or NB-IoT are to be delivered worldwide.21

Sigfox
Scalability
LoRa
NB-IoT
Latency
Range
Performance

Payload
Covertage
Lenght

QoS Deployment

Comparison of LoRa Battery life Cost Efficiency

and NB-IoT in terms
of IoT factors.
(Source: www.sciencedirect.com)

______________________
21
https://www.counterpointresearch.com/5-billion-iot-cellular-module-will-ship-globally-2019-2025-4g-lte-cat-m-nb-iot-hyper-growth/
 Industrial Internet of Things 23

Save on hardware with mesh networks

If IIoT networks become larger, denser and more decentralised, infrastructure costs will
also rise. Mesh networks counteract this effect. The individual IIoT devices function as
satellites in these networks and pass on the data between them to the final recipient
(e.g. via gateway). Due to the lower use of hardware with a high number of nodes, mesh
radio networks are an alternative to the classic star-shaped architecture of local or
wide area networks (LAN / WAN). The software-based Wirepas connectivity protocol is
particularly suitable for setting up networks in the sub-GHz range and 2.4 GHz band.

Local Wide Area Massive IoT

Local areas and small Wide area and sparse Any csale, any density, any
installations - Zigbee, Thread, installations with limeted location installations, OTA, open
BLE Mesh, Z-wave bandwidth - SigFox, LoRa, platform with Wide Area Mesh
Ingenu, NB-IoT

IoT connectivity: development

from a local network to a
Centralized De-centralized randomly scalable Wide
Area Mesh
(Source: Lonmark.org)

Mobile radio networks for global telemetric applications

The advantage of cellular LPWANs, e.g. based on NB-IoT, lies in the higher data rate
and the ability to communicate across long distances and national borders. Compared
to sub-GHz LPWANs, the output power is less limited. The disadvantage lies in the
chargeable use within the framework of a corresponding mobile phone contract. Global
telemetric applications also require international coverage. For example, the Swiss
service provider SAVVY® Telematic Systems AG22 offers a complete end-to-end solu-
tion for worldwide monitoring of tank containers and tank wagons. For international
network coverage by water, rail or road, the Swiss offer their customers an international
all-network subscription at a fixed monthly price. The technical basis for worldwide
connectivity is provided by SIM chips in the telematics devices. The leading providers of
IoT mobile radio modules include SIMcom Wireless, Sierra Wireless and Gemalto.23

SIMcom Wireless 23%

Others 35%

Sierra Wireless 17%

U-blox 5%
Telit 11%
Gemalto 9% Market shares in the global sale
of IoT mobile radio modules
according to unit numbers in
the first half of 2017.
(Source: Q2 2017 IoT Tracker)

______________________
22
www.savvy-telematics.com
23
www.counterpointresearch.com/sierra-wireless-and-gemalto-lead-the-iot-cellular-module-market-in-terms-of-revenue/
24 IIoT GUIDANCE Whitepaper 9/2020

The means of choice for data transmission are the standard protocols OPC UA
and MQTT. IIoT-GUIDANCE collaborates with experienced solution providers such
as SAVVY Telematic System AG to map not only standard-based hardware and
software solutions but also the critical points of scalability, security and future
security.
The technology leader from Switzerland implements the latest security standards
on all levels from the physical sensor over short range radio network and mobile
data network into the cloud and further in the database and visualization in the
mobile device.
SAVVY uses its dedicated secure protocol SSP (SAVVY Synchronisation Protocol) and
has developed a multi-layer security concept consisting of encryption mechanisms
Success factor IT security: that takes into account the entire information chain and includes comprehensive
the multi-layer concept automation mechanisms for scaling.
composed of encryption
mechanisms covers the
entire information chain
(Source: SAVVY Telematic Systems AG)

Logging Logging

User authentication
Backend
Encryption
Wired SAVVY Wireless SAVVY Frontend
Interface Sense Gateway Interface Cargo Trac
API

Transmission
of abstract data Base station
API encryption

Local device access lock Database encryption

Low voltage API (JSON-REST)
external
sensors
ERP
MES

Encryption mechanisms

Level 3 Encryption according customer’s policy

SAVVY

Level 2 Device authentification with changing keys

Level 1 AES encryption Database encryption Encryption according customer’s policy

Level 0 Basic technology encryption

______________________
24
www.savvy-telematics.com
 Industrial Internet of Things 25

Low-power framework from IIoT-GUIDANCE

For IIoT retrofits in the process industry, IIoT-GUIDANCE is developing scalable and pro-
tocol-independent frameworks for the seamless interaction of terminal devices and IT
systems such as SAVVY® Telematic Systems AG technology. The individual components
range from wireless embedded technology to Ex-certified hardware and software to the
easily integrated M2M platform with international data transmission at a uniform rate.

IIoT Cases Frontend & Webservice

Non-Ex
(Rugged)

Div 2 / Zone 2 API

Div 1 / Zone 1
Div 1 / Zone 0

(Source: https://www.isafe-mobile.com / www.savvy-telematics.com)

Summary: success factors of data transmission

A communication approach for seamless interaction of IIoT devices with higher level IT
systems must be convincing in four respects:

1. Efficiency – lean transmission protocol, low data volume, high ranges, energy-
efficient hardware, mesh networks
2. Security – multi-level security concept (data encryption, device and
hardware authentication, see 6.5), multi-level end-to-end encryption, device
virtualisation, certified security through standard protocols
3. Stability – optimised and certified hardware components, high availability:
mirroring of data and software: complete and correct data transfer and
synchronisation, etc.
4. Internationality – international network coverage, worldwide device
management with European or global footprint, direct cost control of devices,
fully automatic and configuration-free data transmission (plug & play), global
and local certificates
5. Future-proof – standard protocols, standard hardware, availability of
components (large, established manufacturers)
26 IIoT GUIDANCE Whitepaper 9/2020

5.3. Communication – networks and protocols

The more work an IIoT device can do, the fewer resources will be needed later in the
cloud. As part of edge processing, collected data is filtered or pre-processed in the
field or at its “edge” (possibly in the secure area). By reducing data accordingly, band-
widths and the energy consumption of battery-operated sensors and devices can be
conserved, but above all short and fast feedback loops can be realised in the process.
In the course of edge processing, the data rate can also be regulated with the aid of
an MQTT broker, so that a transmission takes place, for example, at a defined interval
(e.g. every “x” minute) or only after exceeding or falling below a threshold value. This is
particularly important for image sensors that compare 3D and IR camera images with
reference images (visual threshold) or for vibration analysis. Here, edge processing
can even lead to a self-learning system that has “learned” the basic characteristics
of the still “healthy” pump over a longer period of time and recognises significant
deviations based on the current values. Only then is the data escalated to the cloud for
further analysis. The provider SAVVY ® Telematic Systems AG uses such a procedure
for the worldwide detection of damage to rail freight car wheels.

Local or global data processing?

The degree of edge processing depends on the required degree of accuracy and
up-to-dateness of the data beyond real-time requirements. The aim is to achieve the
desired result with an optimal relationship between local and global data processing
under economic and qualitative criteria. For example, edge processing is the best
method for image recognition, since data transmission is only triggered in the event of
deviations from the reference images. This saves bandwidth and energy in the case of
battery-powered gateways25. In hazardous areas, however, it must also be taken into
account that local computing power is always associated with heat generation, which
Distributed computing power may limit the possibilities of edge processing compared to the safe area. If possible,
for optimum data flow: from the frequent or larger computing tasks should take place outside hazardous areas.
local level via edge processing
to cloud analysis. Gateways can also take over simple edge processing functions. MQTT brokers like
(Source: Knowtion) Mosquitto run under Linux on different platforms like PC, Raspberry Pi or Arduino.

Presentation
Embeded Analytics Edge Analytics Cloud Analytics and decisions

______________________
25
See Chapter 5.4
 Industrial Internet of Things 27

5.4. Gateways – data collection and forwarding

Once the measurement data have been captured, they must be collected and trans-
mitted for evaluation, for example to a cloud or a local server. So-called gateways
implement this task physically and logically. Gateways can display their own hardware
components or run software-based on standard hardware, e.g. an industrial PC. In
order for this key element for the IIoT setup to be usable everywhere, however, the
respective hardware must always be certified for the respective area. There are solu-
tions up to ATEX Zone 1. Data forwarding can be wired or wireless according to require-
ments26. The advantage of a wireless connection is the simple and timely retrofitting,
a shorter project duration and, if necessary, it is possible during operation, i.e. without
shutdown. However, the regular battery replacement must be taken into account, which
can be reflected in the scope of the new IIoT service models if necessary27.

Success criteria for Ex gateways

Equipped with the appropriate computing power, the gateway can also pre-process
or further process data (see 5.3.). The hardware can be flexibly located: while devices
with a large computing capacity are typically housed in the control cabinet, more
compact, battery-powered units are mounted in the immediate vicinity of the data
source. Even lighting components offer a simple and cost-efficient way of accommo-
dating gateways thanks to the existing power supply and can even become part of an
LPWAN mesh28. For wireless connection with wireless sensors, many gateways use the
energy-efficient Bluetooth LE (Low Energy) wireless standard. Smartphones can also
use this standard to connect to the network, for example to make simple parameter
settings. If required, several gateways can also form a short-distance radio mesh
network according to the worldwide standard IEEE 802.15.4, and thus pass on data
over several hundred meters to the next base station29. The protocols for data trans-
port to the cloud server are another success criterion and must be powerful (“rich”)
enough for the respective applications30. This is the case with most industrial proto-
cols as well as with the higher-quality Industrie 4.0 protocols such as OPC UA or MQTT.

Robust and retrofit-capable:

SAVVY® CargoTrac gateway solution
SAVVY®’s battery-powered, easy to retrofit gateway is available for
ATEX zones 1 and 2. The solution developed by SAVVY®Telematic
Systems AG collects data from industrial sensors via an analogue
4-20mA standard interface and wireless with IEEE 802.15.4 or Blue-
tooth LE. Data transmission to the central web interface is via 3G /
4G. The IP67 housing offers chemical resistance and high robustness
for industrial use. The battery life is is up to 15 years. The measuring
points can be managed from any location via the Internet. A multi-
level security mechanism ensures the accuracy of the data and
protects them against unauthorised access.

______________________
26
See Chapter 5.2
27
See Chapter 1
28
See Chapter 5.2
29
See Chapter 5.2
30
See Chapter 5.2
28 IIoT GUIDANCE Whitepaper 9/2020

5.5. Cloud – Use Case and customer specific options

Data processing can be done directly at the sensor, at the gateway level or directly
in the cloud. The cloud is to differentiate between a local cloud, within the corporate
network, self hosted or by an IT provider (even for several warehouses, fabrics) or a
global cloud offered by the hyperscalers e.g. Microsoft Azure, AWS IoT, Google and
IBM Watson, etc.
For big data applications with artificial intelligence (AI)31, there is usually no way
around the large platforms. Computing capacities can be scaled here according to
demand and rented based on usage (pay per use)32. But there are several options pos-
sible, depends on the customer requirements.

The key success factors for the cloud strategy are compliance (IT security, data
access, data protection, etc.) and finally cost-effectiveness, which is derived in par-
ticular from the ability to integrate into existing company applications such as control
centres, SCADA systems or maintenance applications.

Many companies switching from self hosting to IT providers such as Amazon AWS or
Google because of the security aspects they can outsource and the scaling effect
but are not aware of the log-in effects by also using the cloud provider offered instru-
ments. This instruments are great to use but once migated hard to exchange or
replaced by other tools e.g. AI algorithm.

User centric integrated and

holistic IIoT approach.
(Source: IIoT-Guidance GmbH)

Existing Structure IIoT Enhancement by Values

Automation Flow IIoT Flow IT aspects User Groups

Business Models
User
SCADA HMI Mobile Devices Authentification
Device Revenue Sharing
Apps
Analytics Pairing Native Access Control Cost-saving sharing
or
User

Reporting Product-sharing
Cloud Secundary Approval
Optimization Applications
Engineering User Analytics Product as-a-Service
Algorithms
Analysis Performance-as-a-Product
ERP Logging/ Reporting
IIoT Cloud

Transactional

Cloud
Public / Private / Hybrid
MES Operation
Asset & Data Classification
Information Flow

Data Discovery
Data

Gateways Data Exchange

SCADA/ Encryption
HMI
Operating Desk / Control Room
Edge and
Processing
Operator Company

Key Management
CONTROL Local
Service Personal
Infrastructure

Network Authentification
own / external
Configuration Hardening

Local Segmentation
FIELD Sensors Boundary Enforcement
sensor / actuators / controllers Actuators Field Operation
own / external

User centric integrated and holistic IIoT approach

______________________
31
See Chapter 5.6
32
See Chapter 1
33
See Chapter 2
 Industrial Internet of Things 29

5.6. Analysis – big data, algorithms and AI

The purpose of IIoT retrofitting is to obtain information which has not been trans-
parent until now and which can be used to exploit the various potential benefits for
the process industry33. The valuable information gain from the collected and possibly
pre-processed data can be distributed over the three levels - local, edge and cloud -
and differentiates between two basic approaches:

a) Automatic data analysis

In automatic data analysis, sensor data are automatically monitored and deviations
are detected using mathematical algorithms. These algorithms extract the desired
information directly from the data. This includes the entire spectrum of methods from
machine learning, such as linear regression, neural networks, random forest or hidden
Markov models. A typical algorithms pipeline for data-driven approaches that can be
implemented on all three levels consists of the following components:
1. Data pre-processing
2. Feature extraction
3. Feature reduction
4. Actual machine learning algorithm

For this purpose, usually large amounts of measurement data are continuously eval-
uated and interpreted (big data), patterns and deviations are identified and automati-
cally corrected if necessary. The advantages of this procedure are as follows:

ƒ Verification of data in real time

ƒ Early detection of outliers and faults
ƒ Interpretation of data, e.g. process is running as expected / is completed /
deviates from scheduled process, etc.

The typical field of application for automatic data analysis is the monitoring or
adaptive control of processes. For example, “intelligent distilleries” can be realised
that automatically recognise the end of the process. In another example, the sup-
plier Knowtion has developed an automatic monitoring system for the production of
extruded polystyrene rigid foam (XPS). The production process and extruder condition
are evaluated to predict the final quality and properties of the product, and production
is optimised and controlled accordingly.

Prediction
Control
Product quality

Process and
Sensor data
machine data

Adaptive Control
Extruder Intermediate product Final product using automated
data analyses.
(Source: Knowtion).
30 IIoT GUIDANCE Whitepaper 9/2020

b) Sensor fusion
The term of sensor fusion refers to the combination and aggregation of measurements
with several sensors to form a coherent overall picture. These approaches assume
that physical background knowledge is available in the form of a mathematical
description. These so-called model-based approaches now combine the sensor
data with this background knowledge in order to obtain a more accurate result of
the desired information. The best known examples are the Kalman Filter (KF) for
linear systems and Unscented Kalman Filter (UKF), Extended Kalman Filter (EKF) or
particle filter (PF) for non-linear systems. The choice of filter greatly depends on the
application. A typical algorithm pipeline for model-based approaches that can be
implemented on all three levels consists of the following components:

1. Outlier detection
2. Prediction step, and
3. Filter step

By merging different sensor types into a new virtual sensor, it is possible to calculate
data that has not actually been collected. The combination of sensors with mathe-
matical models thus creates added value through additional knowledge based on
correlations34. This can also result in investment cost advantages, since the consoli-
dated overall result of several more cost-effective sensors can be equivalent or better
than the evaluation of fewer, but more precise or expensive special sensors. Further
advantages are as follows:

ƒ Larger amount of information

ƒ Lower investment costs with comparable accuracy
ƒ Improved sensor coverage and reliability by combining different sensors

Virtual sensors are an important area of application for sensor fusion. They are used,
for example, for automatic fault and error detection in the on-load tap-changers of
power transformers, where previously invisible causes such as gas concentrations in
transformer oil become visible. Other typical use cases include the monitoring, recog-
nition and prediction of process and plant conditions.

Automatic detection of the

process condition

Detection of end
of distillation
Algorithm
Pipeline
Intelligent distillery:
automatic regulation of Anomaly and
distillation in the case of fault detection
unknown volume and type
of liquid.
(Source: Knowtion)

______________________
34
See following chapter
 Industrial Internet of Things 31

Temperature

Acceleration

3 State Estimation including Quality Measure

Rotating Rate

Acoustic

2 Knowtion Algorithm 4 Outlier and Anomaly Detection

Humidity

1 Raw Data 5 Short- and Long-term Prediction

Customised analysis Data analysis using

models and algorithms.
(Source: Knowtion).
For automatic data analysis, providers such as Knowtion develop and program tar-
geted software containing algorithms and AI according to certain standards of quailty
and safety. Sensor fusion, on the other hand, is based on mathematical algorithms
and models that use AI and machine learning to obtain higher quality information
from raw data. For this purpose, the experts create various solution approaches within
the framework of an agile process and evaluate them with the help of simulation (e.g.
in Python and Matlab). This allows the following types of use cases to be implemented:

1. Self-monitoring of sensor systems

2. Detect errors
3. Determine / predict sizes
4. Predict failures / errors
5. Track and trace sensors
6. Control processes adaptively (using correlations with peripheral conditions).

______________________
35
www.knowtion.de
32 IIoT GUIDANCE Whitepaper 9/2020

Example: Local anomaly detection on IIoT devices

The algorithms of the model-based and data-driven approaches above can even be
optimised to the extent that they can be implemented on IIoT devices. This will be
shown here as an example for anomaly detection in two different applications: vibra-
tion monitoring of a direct current motor and trajectory monitoring of industrial robots.
The algorithms for both applications are identical, only the parametrisation is differ-
ent, i.e. the time interval considered for vibration monitoring is short and for trajectory
monitoring rather long. The acceleration and gyroscope data with a sampling rate of 1
kHz each were used as input data. For the vibration monitoring, the microphone and
magnetic field data were also used as input data in order to take into account other
conspicuous features such as acoustic changes. This combination of sensors using
intelligent algorithms drastically increased the detection accuracy.

The results of the local calculation on the sensor platform are shown by way of
example in the figures above. Both examples show the acceleration and gyroscope
data, the locally derived characteristics and the locally calculated anomaly indicator.
It can be seen that this indicator increases steeply with new signal behaviour and
is much lower with repeated occurrence, i.e. the newly recognised signal was taken
into account and updated in the model by the learning algorithm. In practice, after
the algorithm has seen all “good data“, a user would stop training to achieve stable
behaviour.”

Vibration monitoring of drives

 Industrial Internet of Things 33

The vision of the self-learning plant

IIoT and artificial intelligence (AI) complement each other perfectly. The IIoT provides
the necessary raw material for the AI with its connected sensor technology and
the data collected using it. For its part, AI can improve the IIoT by evaluating large
amounts of sensor data using machine learning, smoothing measurement errors and
delivering structured data, making data exchange more robust. The ability to create
context, interpret incoming information and trigger appropriate actions can ulti-
mately contribute to the intelligent control and optimisation of the overall system36.
Accordingly, the joint potential of AI and IIoT appears to be many times greater in the
industrial environment than in the consumer market37. An example from the automo-
tive industry illustrates this: the technology behind Tesla Service Autonomous Driving
includes a self-learning system that constantly collects and evaluates data in order
to learn from it and improve itself38. This is based on machine learning algorithms,
wireless vehicle connectivity, and detailed map and sensor data collected by Tesla. The
entire Tesla fleet is always online and works in a network. When one car learns some-
thing, they all learn it because the software is updated “over the air”. In the future, this
could also be used for plants to learn in this way in order to optimise their operation
and share their knowledge in the company network.

In future, self-learning
plants could optimise
operations and share the
existing knowledge in the
company network.

(Source: www.teslarati.com)

______________________
36
https://digitaleweltmagazin.de/2018/02/20/iot-und-ai-eine-co-evolution/
37
https://aibusiness.com/ai-nothing-without-iiot/
38
http://fortune.com/2015/10/16/how-tesla-autopilot-learns/
34 IIoT GUIDANCE Whitepaper 9/2020

Employee empowerment through self-service analytics

The growing range of analytics services can be roughly divided into three groups with
different needs:

1. Generic offers of large IoT platforms such as Microsoft Azure IoT39

or AWS IoT Analytics40 with adaptable default functions
2. Tailor-made solutions from expert firms developing and evaluating
algorithms and models
3. Self-service analytics that empower employees to analyse existing
data and timelines

The large platform providers score particularly well with their randomly scalable com-
puting power and ready-made, easily adaptable analyses and dashboards. In addition,
powerful scripting languages also enable tailor-made analyses. The strength of the
software companies specialising in IIoT analyses lies in their high level of development
and application competence. In both cases, however, a more or less high degree of
internal or external know-how is required in order to develop the appropriate analyses.
This is where an innovative self-service offer from the provider TrendMiner41 comes
in. With easy-to-learn tools, process and asset engineers and managers can carry
out process and asset analyses themselves, take measures, continuously improve
processes and share the acquired knowledge within the company. In this way all oper-
ational stakeholders can contribute to improving operational excellence, reduce costs
and increase overall profitability.

Assets

Customers Process

Actions are Continious

taken to and automated
drive value Field products readings by
sensors

Insights allow Data is

better descisions aggregated
and predictions and processed

Self-service analytics Analytics

empower process and asset reveal trends
Engineers and managers and Patterns
to make continuous
improvements.
(Source: Trendminer)

______________________
39
https://azure.microsoft.com/de-de/overview/iot/industry/process-manufacturing/
40
https://aws.amazon.com/de/iot-analytics/
41
https://www.trendminer.com/
 Industrial Internet of Things 35

Self-service analytics in polymer

production: in this case the
process engineer recognises
the cause of a recurrent
problem by overlapping several
measurement curves.
(Source: Trendminer)

Summary of IIoT analytics – success factors

In the analytics layer, the collected and possibly pre-processed measurement data
are converted into usable information. Consequently, it significantly determines the
return of investment, but is also directly dependent on the quality and reliability of the
data collection. The essential success factor is therefore a well-integrated information
chain and perfect cooperation between the solutions and partners involved - at the
cloud level at the latest. Essential prerequisites for a seamless information chain are
as follows:

ƒ Hardware compatibility (partner and customer systems)

ƒ Seamless collaboration between device suppliers with consulting
companies and system integrators
ƒ Intensive, continuous exchange

IIoT-GUIDANCE is in intensive exchange with the economically and technologically

leading players in the IIoT market. They include device suppliers such as Samson,
consulting companies such as CGI Group42, system integrators (e.g. Augmensys43) and
service providers such as Sitech Services44, Bilfinger Digital Next, SAVVY ® Telematic
Systems AG or TrendMiner – the basis for sustainably successful projects and joint
growth.

______________________
42
www.de.cgi.com/de
43
www.augmented-industries.com/
44
www.sitech.nl/
36 IIoT GUIDANCE Whitepaper 9/2020

5.7. Application level – user devices and apps

Even for the last link in the IIoT information chain, cooperation with experienced part-
ners is essential for the success of the project. In addition to the necessary interfaces
(APIs) for integration into the corporate IT, the visualisation on different HMIs in Ex and
Non-Ex areas also requires corresponding know-how. The success factors here are as
follows:

ƒ High user acceptance

ƒ Minimal IT costs
ƒ Platform-independent display (apps)

As an experienced solution provider, IIoT-GUIDANCE convinces with a sophisticated,

modular and scalable solution concept for visualisation and interaction in hazardous
areas - from stationary PC- and thin client-based HMIs to mobile devices such as
tablet PCs and smartphones with certified accessories. In addition to a platform for
Mobile Device Management (MDM), standardised solutions such as the Tablet PC
series, which - as a uniform platform for Ex and non-Ex areas - minimises project
and IT management costs, are also important for acceptance by the IT department.
IIoT-GUIDANCE also supplies certified automation technology, including HMIs, remote
I/O and bus systems as well as switching and control components, from a single
source for comprehensive digitisation projects45. This overall platform is continuously
extended in collaboration with other leading suppliers.

DIGITAL
PLATFORM

classical end-to-end
picture of a digital platform
(Source: IIoT-GUIDANCE)

______________________
45
www.isafe-mobile.com
 Industrial Internet of Things 37

6 Specific implementation examples

6.1. Beacon/ RHT sensor for the simplification of O&M

Sensor

(Source: IIoT-GUIDANCE)

Sensors send their measurement data via Bluetooth or wired to a gateway. In this way,
previously non-existent data can be collected and made available for use in digital
processes, for example.

Advantages:
ƒ Increased plant availability; no need to open the switch cabinet door or the Ex d
housing (plant shutdown)
ƒ Simple and fast installation / retrofit
ƒ Increased transparency: time stamp for inspections / maintenance (date / time /
person of inspection)
ƒ Storage of temperature and air humidity data (thresholds for falling below/exceed-
ing) and therefore control of the area of application in the event of faults
ƒ Convenient monitoring directly in the customer’s control system (integrated via
standardised interfaces)
ƒ Convenient monitoring, directly in optionally available control system
ƒ Reliable data transmission
38 IIoT GUIDANCE Whitepaper 9/2020

6.2. T
 rack & trace of mobile assets under
Ex conditions
The oil, gas and chemical industries produce and transport huge quantities of valu-
able but also dangerous goods worldwide. Today, companies lack the transparency
in the supply chain to determine where the goods are or where their capital is, and
whether there are transport disruptions or dangerous situations can arise. To create
this transparency, SAVVY ® Telematic Systems AG has developed a standard solution
for the following applications:

The user monitors the process parameters of the product, the plant and the supply
chain in real time - in stationary and especially decentralised plants as well as in
globally mobile containers (rail freight cars, tank containers and containers) - under
the harshest and potentially explosive conditions. This is to ensure the safety and
availability of product and plant at all times.

The user is automatically informed of any deviation of the process parameters from
the expected state. Current and historical values can be viewed worldwide via the
SAVVY ® Synergy Enterprise Portal or can be integrated into the user’s ERP system or
shared with external parties.

The operation of the existing system or the mobile containers is not affected by the
installation of the monitoring solution. The solution can be retrofitted with minimal
intervention in the existing system to keep installation costs and downtimes low and
inspection costs minimal.

The solution is easily expandable and adaptable in order to grow flexibly with the
project or operation of the plant.

The monitoring solution combines sensors with worldwide approved, standardised

IEEE 802.15.4 wireless technology and LTE Cat M1 with NB IoT and 2G backup mobile
network. The solution is therefore no longer limited by wired installations and at the
same time enables increased reliability with low life cycle costs. Critical data is pro-
tected by 128-bit AES security keys.

The following example shows

the fitting of the Ex telematic
device SAVVY® CargoTrac-Ex
to a intermodal waggon of
truck WASCOSA.
 Industrial Internet of Things 39

The following example shows the

monitoring of a container on the
new Silk Road between Europe and
China, where impacts, downtimes and
schedule deviations are monitored.
(Source: SAVVY® Telematic Systems AG)

The SAVVY Telematics® CargoTrac-ExR-M1 telematics device is allowed to travel in

danger zones with potentially explosive atmospheres, because the telematics device
is approved in accordance with Directive 2014/34/EU for potentially explosive areas in
zones 1 (gas group IIC) and 21 (dust group IIIC) and is also IECEx-certified.

Thanks to its highly efficient battery supply, the monitoring solution is completely
self-sufficient and maintenance-free, with the exception of battery replacement after
10 to 15 years, for example.

Advantages:
ƒ Transparency in the supply chain for the consignor and the consignee as well as for
all service providers involved in transportation
ƒ Estimated time of arrival (ETA) determination
ƒ Hazard detection thanks to sensor technology
ƒ Predictive maintenance thanks to on-board vibration analysis
ƒ Localisation of damage to product or container
ƒ Increased plant and container availability
ƒ Reduction of failures thanks to immediate intervention options
ƒ Convenient monitoring directly in the customer’s control system (integrated via
standardised interfaces)
ƒ Convenient monitoring directly in the SAVVY ® Synergy Enterprise web portal
ƒ Reliable and secure global data transmission
ƒ Simple and fast installation / retrofit even during operation
40 IIoT GUIDANCE Whitepaper 9/2020

6.3. Valve position monitoring

Fusion sensor

Gateway

(Source: IIoT-GUIDANCE)

There is a large number of manual valves in process plants; many existing plants have
been working successfully and effectively with them for a long time. Valve positions
are increasingly monitored in order to optimise process operation and prevent oper-
ating errors. For this purpose, a valve position sensor is attached to the valve. In the
event of a position change, the sensor sends a signal to a gateway, which transmits the
information to the control room, where the personnel can oversee the valve positions.

Advantages:
ƒ Greater plant availability due to continuous valve position monitoring
ƒ Reduction of failures thanks to immediate intervention options
ƒ Convenient monitoring of manual plant parts directly in the control system
ƒ Reliable data transmission
ƒ Simple and fast installation / retrofit
 Industrial Internet of Things 41

6.4. Corrosion monitoring of pipelines

2 km

Gateway

(Source: IIoT-GUIDANCE)

The corrosion monitoring of pipelines for perfect function is an essential prerequisite

for the availability, performance and safety of plants. The quality management of
these pipe systems includes a thorough analysis of the actual condition. This is done
today by visual inspection and mainly during plant shutdowns. Digital monitoring
holds great potential for service companies and ultimately for the operator / owner.

For this purpose, sensors are installed between the insulation and the pipeline. They
send the position / temperature and humidity to a gateway via Bluetooth. Here, the
data is collected and sent to the control room or to an evaluation system in the cloud.
Using suitable algorithms, the corrosion can now be viewed as a function of ambient
temperature and process.

Advantages:
ƒ Greater plant availability through continuous monitoring
ƒ Reduction of failures thanks to immediate intervention options (transparency)
ƒ Savings on manual, expensive tests through automatic measurements
ƒ Convenient monitoring of pipeline parts directly in the control system
ƒ Reliable data transmission
ƒ Simple and fast installation / retrofit
42 IIoT GUIDANCE Whitepaper 9/2020

Barriers and success factors in

7
IIoT retrofitting in hazardous areas
Security, IT/OT integration and unclear ROI: these are the three points that Bain &
Company’s customers cite as the biggest obstacles to introducing the IoT46. This reflects
the high priority the market places on securing and integrating IoT networks. Compared
to the previous survey, the importance of interoperability, data transferability, vendor
risk and network constraints has also increased.

Vendors needs to address customer barriers - especially security,

integration and unclear returns on investment

What are the most significant barriers limiting

your adoption of IoT/analytics solutions? Change since 2016

PERCENTAGE OF RESPONDENTS (TOP THREE BARRIERS)

Security
IT/OT integration
Unclear ROI
Tecnical expertise
Interoperability
Data portability
Vendor risk
Transition risk
Legal/regulatory issues
Network constraints
Vendor lock-in
Customer survey: the most
significant barriers limiting
adoption of the IoT. 0 10 20 30 40 50
(Source: Bain & Company)

The business consultancy therefore recommends that providers narrow their focus
to a few target groups. With its eco-system partners, IIoT-GUIDANCE summarises
the most important success factors to significantly simplify the complex task of
setting up IIoT are as follows:

______________________
46
www.bain.com/insights/unlocking-opportunities-in-the-internet-of-things/
 Industrial Internet of Things 43

Application examples

Technical success factors Process monitoring, alarm on exceeding or falling below a threshold
value, monitoring of overland lines

Process security Terminal devices and IT systems interact seamlessly

IT security Stringent, multi-layer IT security concept: encryption, data protection

(information in containers), role-based data access, etc47.

Economic efficiency Energy-efficient design of all partial solutions (long battery life, low
maintenance); efficient network infrastructure (minimal use of hardware,
e.g. by mesh networks); sensible use of edge processing (transmission
only in case of deviations); continuous improvement through self-
learning systems, etc.).

Simplicity Reduce complexity through sensors that can be retrofitted easily,

efficient networks, device management, automatic software updates,
efficient analysis development, user and IT friendly modular solution
design for visualisation and interaction with the user

Integration Interfaces to existing IT systems and partners (e.g. tracking warnings

and forwarding them to maintenance partners); business intelligence
tools for setting up customer-specific dashboards, mobile apps for fault
documentation, etc.

Future-proof The complete solution must take into account the complete product life
cycle of all individual components. Independence of the overall solution
from communication protocols (see 5.2.) Size of the producer, availability
of components, future prospects for implementing protocols

Scalability When selecting solutions, pay attention to scalability, e.g. device

management with automated SIM management and software updates.

Organisational success factors

Human factor: know how transfer Transfer of employee know how into digital processes. Constructive
and empowerment handling of scepticism towards innovations. Opportunity to replace
simple activities with knowledge work. Empowerment of employees
to actively improve processes: e.g. through self-service analytics,
thereby also sharing know how in the company or across locations;
direct monetary participation of employees in improvement processes;
transparency in the field through enterprise mobility

Access to Ex know how Selection of an experienced partner with holistic solution approach:

- mobile Ex platform with option for data acquisition and integration of

existing information channels

- comprehensive range of ATEX-certified terminal devices and

accessories

Start small, improve quickly In reasonable steps from the small to the big solution. First small use
cases with fast results (ROI < 1 year), then further use cases. End-to-
end solutions with concrete benefits.
Project support Experienced partner to support and moderate between departments
and solution partners

______________________
47
See Chapter 4
44

Checklist:
how to incorporate IIoT solutions?

 o you have enough ressources to start projects parallel to the daily

 D
business?

 G
 et a experienced partner on board. Learn from failures of others
and do not reinvent the wheel!

 U
 se our Approach and Solutions forms for a structured and strategic
approach. Additional we will guide you with our custom-tailored IIoT
workshop.

 I dentify an initial use case. Focus on the right sensor concept (the
right sensor will be the fundament of the Use Case) with RoI < 1 year
to get results soon.This is important for team motivation.

 T
 echnically discuss the use case in detail as described in our
Solutions form.

 D
 ecide for a final communication protocol and connectivity option
(continious costs vs. one time costs vs. service over the life time
vs. ownership of the infrastructure vs. traffic costs vs. need of
bandwith). (see chapter 5.2.)

 T
 hink on a service partner plan (know how/ long-term cooperation/
repair/ service and battery exchange for e.g. wireless sensors).

 C
 reate acceptance within the company. Involve C-level and affected
departments at an early stage (IT, service, partner sales, etc. with
IIoT-GUIDANCE as moderator).

 C
 larify target groups internally and externally including price and
billing model (CAPEX -> OPEX).

 H
 ow the data should be accessable (e.g. dashboard vs. mobile APP
vs. integration of the existion ERP eg. SAP).

 W
 hen developing the technology concept, pay attention to
compliance, practicability (possibly during operation) and future
security (energy supply, communication solution, edge computing,
automation, cloud platform, IT security, data protection, etc.).

 I mplement a pilot project (possibly during ongoing operation)

with RoI < 1 year

 Recap and plan further steps and projects

 F
 or a global rollout think on local certificates and approvals (check
chapter 3)
 Industrial Internet of Things 45

8 Summary and conclusion

The Industrial Internet of Things is now also about to make a breakthrough

in the process industry. In combination with various analysis options,
retrofit-capable solutions release enormous optimisation potential and
open up completely new service and business models in order to reduce
CAPEX. The importance of software is increasing significantly; former
hardware providers are becoming brokers and service providers. With the
availability of infrastructure even for hazardous areas and new technologies
for energy-efficient data collection, processing and transmission, the last
hurdles to economical retrofitting are falling. In combination with AI as a
“game changer”, the IIoT will become the key technology for continuous
and economic process improvement in the long term.

The greatest benefits are promised by holistic solution approaches, which

are developed together with experienced solution partners. As a partner and
enabler for the introduction of IIoT solutions, IIoT-GUIDANCE is developing
a complete and scalable “ECO system” according to the requirements of our
customers.

Anyone wishing to fully exploit the possibilities of the IIoT must think
holistically and strategically, but also take their first small steps. Those
who miss the start will have a hard time catching up with the knowledge
lead of competitors.

Would you like to be at the forefront? Find out how you can increase
efficiency, transparency and thus security in your company?

Launch your first IIoT project now with IIoT-GUIDANCE at your side!

Digitisation is an opportunity,
not a threat. Hazardous areas
are no longer show stoppers.
Take the first steps now with
IIoT-GUIDANCE at your side!
Points of contact:

Mirko Lampe
Founder
Yorckstrasse 6, 14467 Potsdam, Germany
Mobile: +49 152 584 388 94
[email protected]

Vesa Klumpp
Managing Director
Knowtion UG, An der RaumFabrik 33c, 76227 Karlsruhe, Germany
Phone: +49 721 486 995-12
[email protected]

Felix Schwarz
Head of Business Development
SAVVY® Telematic Systems AG, Grabenstrasse 9, 8200 Schaffhausen, Switzerland
Phone: +41 52 633 46 05
[email protected]

Thomas Dhollander
CTO and Co-founder
TrendMiner NV, Kempische Steenweg 309/5, Corda Building 2, 3500 Hasselt, Belgium
Phone: +32 484 778001
[email protected]

Benedikt Eckert
Head of partner business development
i_Park Tauberfranken 10, D-97922 Lauda-Königshofen, Germany
Phone: +32 484 778001
[email protected]

iiotguidance.com