Data
Management
for industrial
machines and
plants
breton white paper
Breton©2023
Milan Federico
Software Product Manager
Breton Digital Hub
Data Management for industrial machines and plants
Contents
Contents
Breton presents the second in a series of
White Papers – informative documents
intended to promote and highlight the
characteristics of various topics.
This White Paper addresses the importance of
data-drivenness, construed as the pursuit of a
data-based decisional process.
Information can be used to improve corporate
processes and achieve a competitive edge.
Data is important because it allows us to work
more efficiently, optimising production, and
can also help in making a process sustainable.
Current situation
Page 4
Customer needs
and new trends
Page 14
Proposal of
Breton software
solutions
Page 20
Data-drivenness
Page 10
Objectives
achieved
Page 18
Book a demo
Page 24
Data Management for industrial machines and plants
Introduction
Process manufacturing embraces a vast
range of sectors, such as the production
of consumer goods, chemicals, food and
pharma, all of which use machines and
systems to transform raw materials into
finished products.
In many cases the industries concerned
use advanced technology to automate and
optimise their production processes, but
sometimes the machines they use are not
connected and therefore prevent the data
generated during the process from being
collected and used. This can be problematic,
because data collected by machines during
the production process can be used to
optimise processes, boost efficiency, and
reduce waste.
Data collection in real time can help identify
and solve problems fast, thus preventing
production downtimes and reducing costs.
In addition, the use of data to optimise
processes can help to make the manufacturing
industry more sustainable, reducing its
environmental impact and increasing energy
efficiency. Without being able to collect and
use data, process manufacturing companies
simply cannot fully exploit these advantages.
To overcome this problem, some industries
are starting to invest in connectivity solutions
for their machines and equipment so they can
collect and use their data in real time in order
to optimise their production processes and
make the industry more sustainable. However,
this often calls for substantial investments in
technologies such as sensors, connectivity
gateways and data analysis software, and it
may be some time before any meaningful ROI
can be expected.
Introduction
Data Management for industrial machines and plants
current situation
Many industries are
working to improve their
global competitiveness by
exploiting the data
in their possession to
its full advantage.
1 / Current situation
Data Management for industrial machines and plants
1 / Current situation
Until now the market has offered many
solutions in the realm of data-driven
technology, many of which verticalised and
sector-specific.
The current offering however provides at best
only a partial response to the demands and
requirements of customers.
Companies in the manufacturing sector
are thus left with unmet needs and so risk
adopting a series of mutually-incompatible
micro-solutions.
We also see a proliferation of technologies
that frequently fail to give the hoped-for
results due to additional inefficiencies in terms
of time, organisation, and cost.
So, apart from acquiring multiple suppliers
there is also the cost implicit in understanding
how to make them work together.
There are many information-rich processes in
the manufacturing sector, and companies are
keen to collect data in order to monetise it and
use it to their advantage.
Context
Data Management for industrial machines and plants
1 / Current situation
“What we have
is a data glut.”
Vernor Vinge
Science Fiction Author
Professor of Mathematics and Computer Science
Data Management for industrial machines and plants
1 / Current situation
Excessive data volumes can have a negative
impact on a company's business.
For example, if a machine reads a temperature
at 10 second intervals, the result is 8,640
samples/day and 3,153,600/year.
What to do when you have to manage so
much data? Who uses them and why?
Collecting data has a cost! And the cost is high!
If we need to collect temperature data, we
need a sensor and a connection, but first of all
we need an electronics engineer to establish
where to install the sensor and where to
make the connections, not to mention cable
routing requirements.
We’ll then need an acquisition board but that’s
Problem: Data overload
The information contained within
the data do not allow value to be
imparted unless the data is first
managed, filtered, and analysed. It
follows that data is valuable not in
itself but only if it is used properly
and contextualised.
not enough, because we’ll need a software
architect to write a suitable program. Now
we only have a value, such as a value per
second... how use it?
This is where it gets difficult: each item of data
must be historicized, given meaning,
and used.
The data must therefore be useful, able to
impart value.
Data Management for industrial machines and plants
1 / Current situation
Production machines must produce. From the
perspective of the production plan (parts on
the Y axis, time on the X axis), production has
a simple formula:
N = (Parts per hour) * Time.
The more I want to produce (in the same time
period) the faster I need to work.
If we add another dimension to our analysis,
such as the energy axis, we find that energy
proceeds with speed squared.
Why do we need to change the paradigm?
“We’re entering a new world in which data may
be more important than software.” Tim O’Reilly
If correctly collected and used, data can
show, for example, that manufacturing rapidly
means increasing consumption levels. To
contextualise this concept in everyday terms,
if I drive at 100 km/h I will use far more fuel
than when travelling at 70 km/h: I follow the
same roads but in less time and at far higher
cost. Where is the sweet spot?
Data Management for industrial machines and plants
If we add another axis to our analysis, such
as the average life of machine components,
we discover that the faster we go, the higher
the stress on the machine and the faster its
components wear out.
A data-driven approach makes it possible
to find sweet spots, selecting the optimal
operating areas. If we add an axis showing
“energy cost over 24 hours”, we might
find an advantage in choosing night-time
production, for example, and the speed and
machine lifetime data may justify a drop in
hourly performance figures while maintaining
daily output unchanged and reducing costs
during the night hours by reducing machine
performance levels.
The concepts of data-drivenness thus make
it possible to apply optimisation strategies
in terms of energy, quality and process.
Productivity is no longer a linear formula,
and optimisation is possible only through the
efficient use of data.
The aims of the data-driven mindset are to
optimise, rationalise, and improve, and they
bring also concepts of sustainability.
The ability to predict future events allows
companies to automate, work longer in
unattended conditions, or reduce stress
affecting operators. In short: increase in
quality and decrease in costs (once hidden,
but today visible in sharp relief).
1 / Current situation
Data Management for industrial machines and plants
10
DATA-DRIVENness
Data-driven logic
combines a vast amount of
data and information,
merged and selected to
obtain an appropriate level
of quality and integrity of
the information that will be
used for decision-making.
2 / Data-drivenness