JEC COMPOSITES MAGAZINE - Issue #119 - Supplement lnnovation report - 5

Paris, France
March 6-7-8, 2018

Composite part manufactures have already
recognized that manual processes are prone to
human errors, meaning that operators' understanding of events on the production floor is
usually limited to those which they can physically see and hear in hindsight. Moreover,
lack of contextual information can result in
production delays that impact the entire production floor, often leading to increased costs
and reduced efficiency, raising the prospect of
inferior quality products or missed delivery
deadlines.
Consequently, many composites manufacturers have found that comprehensive asset
tracking capabilities allow the elimination
or improvement of many manual processes.
However gaining full traceability generates
massive amounts of data, which present the
need for efficient data management and smart
analysis.
Cloud computing meets manufacturing
When a sensor network has been widely
deployed across a factory, it tracks events in
unprecedented detail. Over time, this data
delivers full traceability - the digital thread
that maps the lifecycle of a product from raw
material, through the manufacturing process
and beyond to MRO (Maintenance, Repair
and Operations). Managing this data presents challenges and when IIoT solutions are
widely deployed across a production site, the
datasets being collected are often unimaginably large.
Storing such quantities of data on-site could
quickly cause capacity problems, leading to
a need for major infrastructural investments.
The high cost, and the need for ongoing maintenance of on-site data storage systems of the
size required is impossible for most manufactures. With the cloud however, storage space
is unlimited and when factories need more
space, the system is instantly scalable.
Given the vast quantities of data that are now
being gathered round-the-clock in composites factories, storing it in on-site servers is
rarely an option. Instead, manufacturers have
opted to store it off-site, though securely, in
the cloud. A key benefit of the cloud is security while firms such as 'Microsoft Azure' and
'Amazon Web Services' are more than experts

at protecting their customers' data. Contrary
to popular assumptions, cloud computing is
also low cost. Data storage facilities must be
secure, they must be low cost, they must provide capacity for massive and ever-growing
data-sets and, crucially, they must provide
the immense computing power necessary to
analyze substantial amounts of data in a useful
and efficient fashion.
AI and the Digital Assistant
The application of AI-based software allows
the full potential of IIoT to be unleashed.
The AI-based software acts as a 'digital assistant' to factory staff, delivering the alerts and
recommendations derived from the insights,
and allowing managers and operators to enjoy the full benefits of IIoT. An AI-powered
digital assistant can 'see', monitor and 'understand' hundreds or thousands of variables in
real-time, using this information to make recommendations for human operators, or even
act entirely autonomously on their behalf to
optimize production. This represents a radical
change in how factories are run, automating
and optimizing the entire production process
and making the factories 'smart'.
IIoT together with cloud computing and AI
algorithms all come together today, to offer
manufacturing managers a true digital assistant that can eliminate error-prone and
inefficient manual processes. These digital
assistants can communicate with factory floor
workers on-the-go via mobile devices, giving
every worker a contextual awareness that extends beyond their own human capabilities,

Digital assistants collect data from factory floor
sensors to provide complete visibility across inventory, work orders, machines and tools

and when needed - across the entire business.
This contextual awareness can be applied to
improve productivity in any stage of a composite manufacturing facility and across multiple sites.
This raises the question of how best to implement IIoT technology in the digital factory.
Three examples of implementation processes
that are often used are optimized material selection, tool maintenance, and quality control.
Optimized material selection
with the Digital Assistant
The method used to select freezer-stored carbon fiber raw material has a critical impact on
efficiency. Among the many parameters considered, are raw material exposure time left
(ETL), and roll length. Yet factory staff rarely
have the time or ability to study all the variables, and at times, crucial information is often
unavailable at hand, and therefore not taken
into account. Understandibly, factory staff
tend to grab the nearest raw material roll they
find, trying to spend less time in the freezer
and hoping to work quickly to improve productivity. This manual and random material
selection process has large repercussions on
waste and cost.
An IIoT/AI-based solution for selection
of time-sensitive materials can be highly
effective. It monitors every roll of raw material equipped with an RFID or BLE tag.
Compatible readers placed around the production floor scan material rolls as they enter
the manufacturing facility, allowing crucial
information, such as ETL and roll length to
be stored in a secure, off-site cloud. When an
AI-powered digital assistant has access to all
this information, it analyzes it instantly and
uses it to select a material roll of optimum
length and with the shortest ETL.
Workers responsible for removing rolls from
the freezer can be equipped with hand-held
mobile devices so that the digital assistant
can 'communicate' with them, directing
them to the optimal roll selection each time.
By considering more information than a human mind could ever process and analyzing
it in a way that the human mind is simply
incapable of, this process has the ability to
eliminate human errors.

N° 119 supplement - Innovation Report

/ jec composites magazine

Table of Contents for the Digital Edition of JEC COMPOSITES MAGAZINE - Issue #119 - Supplement lnnovation report