Part 2

Cover Story

Data Brokers Foster

IT/OT Convergence
and Interoperability
Obtaining the most value from data requires effective communications among industrial internet
of things (IIoT)-connected devices that allow integration between operational technology (OT)
and information technology (IT) systems. Data brokers play a key role in allowing communication
between the two areas

D
Ravi
ata sets have become a critical technology (IT) systems by efficiently bridg- Subramanyan
asset at chemical manufacturing ing the two areas. Industrial internet of things HiveMQ
facilities, and is the currency that (IIoT) tools and data brokers play a very im-
enables chemical process indus- portant role in ensuring that the data are
tries (CPI) companies to benefit from Indus- available for advanced use cases that allow
try 4.0 technology. In most digitalization ap- chemical manufacturing organizations to
plications, data originate from controllers, fully benefit from Industry 4.0 technology.
programmable logic controllers (PLCs), This article discusses the role and operation IN BRIEF
gateways and edge-computing devices, of data brokers in IT-OT integration. DATA MATURITY MODEL
and move toward advanced manufacturing IT-OT COMMUNICATIONS
technologies, such as artificial intelligence Data maturity model
(AI), digital twins, augmented reality (AR) Data have become an essential component DATA BROKERS
and others. This flow of data enables in- of chemical manufacturing as companies CONCLUDING REMARKS
tegrated, autonomous and self-organizing work to digitally transform and adopt In-
manufacturing systems. dustry 4.0 tools. Chemical companies are
In order to convert data into plant and generating vast amounts of data every day,
process wisdom, there are several key chal- from laboratory experiments to supply-chain
lenges that need to be addressed. One sig- management. These data have the potential
nificant challenge is integrating operational to help chemical manufacturers make better-
technology (OT) systems with information informed decisions, optimize processes and

FIGURE 1. Digital maturity is indicated by the ability to transform raw data collected from a device into actionable information
for decision makers

CHEMICAL ENGINEERING WWW.CHEMENGONLINE.COM JUNE 2023 27

 The journey from data collection ufacturing plant are heterogeneous
to digital maturity in chemical manu- and use various protocols to com-
facturing is one in which analysis, municate. Data generated during
context and insights are added to chemical manufacturing can vary
transform raw data captured from significantly in terms of format, qual-
a device or system into information, ity and completeness. This can make
knowledge, and finally, actionable it difficult to collect, integrate, and
wisdom for decision makers in the analyze data effectively.
facility (Figure 1). Data connectivity is also a major
First, data are collected from issue due to the archaic, legacy na-
chemical manufacturing processes ture of factory systems. As a result,
FIGURE 2. Information technology (IT) and opera- and equipment, such as boilers, heat typically IT and OT systems do not
tional technnology (OT) systems need to communi- exchangers, tanks, furnaces and so have an easy way to communicate
cate easiliy to enable Industry-4.0 initiatives
on. They are then normalized, digi- to enable Industry 4.0 initiatives
drive business growth. tized, and organized as “big data.” (Figure 2).
In addition, the chemical manufac- Next, meaning is added and data IIoT, a subset of Industry 4.0, uses
turing industry is required to comply are synthesized into knowledge via smart sensors and actuators along
with a complex set of government AI. Finally, the data are transformed with software to consolidate data to
regulations. An effective compliance into actionable wisdom attained enhance manufacturing and indus-
program is essential to prevent busi- through the combined insights of trial processes. IIoT is a key enabler
ness disruptions and loss of reputa- digital maturity. to achieving IT-OT convergence,
tion. As a result, chemical manufac- and the interoperability of the vari-
turers must maintain strict quality IT-OT communication ous communication protocols. It
control, detailed product information, The first and most significant frontier does it by creating a data abstrac-
and a continuous integration of the to achieve digital maturity for Industry tion layer in the middle and a com-
data between OT and IT systems. 4.0 is data collection and data move- mon data language to translate the
Due to the drive toward digital ment. Data from chemical manufac- various communication protocols to
transformation, plus the need for turing machines, processes, and en- enable interoperability.
regulatory compliance and quality vironment are captured via sensors This IT-OT convergence is impor-
control, data and analytics capabili- and stored via several key technolo- tant because success in today’s con-
ties in the chemical manufacturing gies. On the OT side, data are stored nected industrial landscape hinges
area have leapt forward in recent with controllers, PLCs, gateways and on collaboration. IIoT is changing
years. The volume of available data edge devices, and on the IT side with how manufacturers work, blurring
is growing exponentially, and more a data center or enterprise cloud. the lines between IT and opera-
sophisticated algorithms have been Data storage technology enables the tions. For example, IT professionals
developed and computational power long-term storage of digitized data might now spend more time working
and storage have steadily improved. captured from advanced sensors. with equipment on the factory floor,
Therefore, instead of intuition, the This data-rich environment enables while OT teams will need to focus
new normal is to rely on data to drive advanced initiatives such as ML, AI, on cybersecurity and networking
digital innovations and business de- adaptive control and digital twins. best practices. IT-OT convergence
cisions. Indeed, data are the “most There are some challenges to data isn’t about turning IT professionals
valuable resource” for chemical man- collection and data movement in into heavy machinery operators or
ufacturing organizations. chemical manufacturing. Machines plant engineers into data scientists.
Industry 4.0 brings together ad- and processes in the chemical man- Rather, it is about creating a strategy
vanced manufacturing technologies
like AI, machine learning (ML), digital
twins, AR and virtual reality (VR) to
enable integrated, autonomous and
self-organizing manufacturing sys-
tems that operate independent of
human intervention. Chemical man-
ufacturing machine and process
data can be analyzed by algorithms
and used for critical real-time busi-
ness and operational decisions that
directly help achieve higher overall
equipment efficiency (OEE), enable
remote monitoring, reduce carbon
footprint and enable product inno- FIGURE 3. Data brokers are intermediary entities that enable OT annd IT cliient systems to communicate
vations. with each other

28 CHEMICAL ENGINEERING WWW.CHEMENGONLINE.COM JUNE 2023

that bridges the gap and allows or- the above-mentioned reasons. the broker. The SCADA client would
ganizations to improve operational The MQTT data broker is able to publish data to the broker when
performance by working around a securely communicate the data be- available. As a result, the streaming
unified set of objectives and key per- tween publishing clients typically on analytics application subscribed to
formance indicators (KPIs). the OT side to subscribing clients on the broker would automatically re-
the IT side. For example, a stream- ceive the updates without needing
Data brokers ing analytics application might want to poll for the data.
A key technology enabler for IIoT to data from the SCADA system to run MQTT technology is designed to
facilitate this data abstraction layer its analytics and publish real-time re- push data to and from thousands
is a data broker. A data broker is sults. The application would run an of remote devices across numerous
an intermediary entity that enables MQTT client that is subscribed to sites to the enterprise. Sparkplug
OT and IT client systems to com-
municate with each other. Using an
underlying standard, such as MQTT
(message queueing telemetry trans-
port), a data broker supports the abil-
ity to have multiple clients connected

A data broker is an
intermediary entity that
enables OT and IT client
systems to communicate
with each other.

that are publishing data and multiple

clients that are subscribed to receive
the data, such as enterprise applica-
tions (Figure 3). The clients commu-
nicating with the broker can abstract
the underlying protocol that the ma-
chines and processes use to com-
municate. The broker works well in
low bandwidth environments with
unreliable communication mecha-
nisms due to the underlying publish/
subscribe method where machines
and processes do not need to keep
polling to get the data.
MQTT is an open-source mes-
saging protocol for machine-to-
machine communications. It is a
standard binary publish-subscribe
messaging protocol designed for
fast and reliable data transport be-
tween devices especially under very
constrained conditions. Constraints
include unreliable network con-
nectivity, limited bandwidth, limited
battery power and so on. It is built
on top of TCP/IP which is the go-to
communication protocol to inter-
connect network devices on the in-
ternet. MQTT is ideal for IIoT due to For details visit adlinks.chemengonline.com/84648-13
CE Island Half Page Ad – Water Footprint – 100 MW Plant.indd 1 12/20/22 3:49 PM
CHEMICAL ENGINEERING WWW.CHEMENGONLINE.COM JUNE 2023 29
 materials as they transform into a
sellable product. It may not be re-
quired, but it can put chemical man-
ufacturers ahead of the competi-
tion by building a better relationship
between a business and its clients
during each step of the supply chain
process. An MQTT-based data bro-
ker enables this product traceabil-
ity use case by ensuring that data
collected by PLCs at the edge from
boilers, furnaces, and other devices,
is combined with product quality
and inspection records coming from
other applications, like enterprise
FIGURE 4. In an MQTT Sparkplug-based data architecture, the data broker connects multiple machines
to enable bidirectional data movement resource planning (ERP) and manu-
facturing execution systems (MES).
The result is a single source of truth
for all of the chemical manufacturing
product traceability data to enable
better product quality and increase
in profit margins.

Concluding remarks
When harnessed correctly, data are
the most valuable asset for many or-
ganizations, particularly in chemical
manufacturing. In order to achieve
the most benefit from Industry 4.0
technology, data need to be trans-
formed to wisdom. A key initial step
FIGURE 5. The diagram shows how data collection and movement could be managed at a chemical toward this journey is data collec-
manufacturing facility
tion and movement. When it comes
is a framework that sits on top of seamless bidirectional data move- to bridging data efficiently from OT
MQTT to add more context to the ment. This architecture solves the systems to IT systems, IIoT and data
manufacturing data. It is an open- chemical manufacturing challenges brokers play a key role to ensure that
source software specification that of data integration and real-time data the data are available for advanced
provides MQTT clients with a frame- movement while providing efficient use cases. MQTT and Sparkplug
work to integrate data and provide and secure data transfer to support technology based solutions can help
context by defining data models. It digital transformation initiatives. users solve their data collection and
provides a consistent way for equip- Figure 5 shows an example of how movement problem, allowing them
ment manufacturers and software data collection and movement is to focus on the advanced data-ana-
providers to share contextual data, managed in a chemical manufactur- lytics use cases to optimize their fac-
accelerating the digital transforma- ing plant to enable a product trace- tory operations. n
tion of existing operations. ability use case. For any chemical Edited by Scott Jenkins
Sparkplug allows IIoT deployments manufacturing firm to maximize
to decouple the data between hard- profit margins, it is crucial to have Author
ware and software sources. With adequate controls over the process, Ravi Subramanyan is the direc-
Sparkplug, new data sources are im- quality, and delivery of its products. tor of industry solutions manufac-
turing at HiveMQ (Postplatz 397,
mediately discoverable to other sys- Traceability programs can play an 84028 Landshut, Germany and
tem components and these sources integral part in this process by lo- Registry Court Landshut HRB 890;
can become a single source of truth. cating problems (and potential is- Phone: +49 1746021120; Email:
[email protected]).
Sparkplug is fully secured, requiring sues) with product quality at the Subramanyan is a product market-
no open ports for new devices and earliest stage possible. Chemical ing and management leader with
extensive experience delivering
requiring TLS for all data transport. manufacturers are often concerned high-quality products and services that have generated
Figure 4 provides an MQTT Spark- about ensuring product quality and revenues and cost savings of over $10 billion for com-
plug-based data architecture that safety for the final product delivered panies such as Motorola, GE, Bosch and Weir. Subra-
manyan has successfully launched products, estab-
shows how the data broker con- to customers. However, most over- lished branding and created product advertisements
nects multiple machines and pro- look “outbound traceability” — the and marketing campaigns for global and regional busi-
cesses, and applications to enable complex process of monitoring raw ness teams.

30 CHEMICAL ENGINEERING WWW.CHEMENGONLINE.COM JUNE 2023