Efficient Plant February 2020 - 13

feature | hmi technology
End users need analytics to improve systems and
advanced HMIs can be used to retrieve necessary
data, regardless of the source. HMI configuration
for this data-centric model should be as intuitive
as it would be if the data was coming from a single
local device such as a PLC. Courtesy of ADISRA

DRIVING DATA HANDLING

One option for HMI data handling is to
collect and store all possible raw data. Users
may not know at configuration time just
what data will be relevant for later analytics,
so future needs can be addressed by storing
and forwarding all collected raw data to
a local database or to the cloud. Future
machine learning and artificial intelligence
initiatives will require massive amounts of
raw data to create insights. Carefully
designed HMIs can still handle this raw
data, but store it and forward it in a lessthan-real-time process to balance networking demands.
Sometimes it isn't appropriate to transmit all possible connected and derived
data values to higher-level systems at full
fidelity. Loading up networks and consuming expensive historization licenses
and server resources doesn't always make
sense. In these cases, a good solution is to
pre-process the data in the HMI so only
the useful results are communicated. This
pre-processing could include level filtering,
totalizing, reporting upstream by exception,
and creating derived values such as overall
equipment effectiveness (OEE).

NEXT-LEVEL HMIS
Today's HMIs are changing the experience
of developers and end users in several

FEBRUARY 2020

EP2002fadisra.indd 13

ways. New IT-capable and cloud-friendly
protocols such as MQTT and AMQP are
readily incorporated in HMIs, allowing
them to reliably communicate with supervisory systems, even through intermittent or
low-bandwidth connections.
It is important to note that an HMI can
be a hardware/firmware combination with
local connectivity, or a software solution
running on various levels of hardware with
diverse types of connectivity. HMIs can be
hosted on PCs, edge devices, centralized
servers, or hybrid combinations of these-
providing modularity and design flexibility.
The underlying operating system (OS)
for any given HMI application may vary as
an HMI's host OS is no longer limited to
Microsoft Windows. Open source OSs, such
as Linux, are rapidly becoming acceptable in
the industrial environment.
While HMIs have traditionally been
local, the ability to locate them anywhere
and push data to any cloud service such as
Microsoft Azure or Amazon Web Services
is not only possible but expected. Modern
HMIs also use common interface platforms
such as HTML5 to generate graphics and
interface pages. Using these platforms, users
will be able to view and interact with their
processes and data anywhere with mobile
devices, such as smartphones.

Most important in our global environment
is visibility into the process and data-data
anywhere, anytime. Today's HMI can meet
these needs and more. Imagine an OEM
deploying fleets of machines with global
HMI access to provide support, determine
OEE, and address maintenance issues. This
type of service model will become commonplace as a result of the evolution of
HMI software.
A modern-day HMI will still look and
feel much like the HMIs of recent years,
although there will be significant additions
under the hood. The continuing advancement of HMI software applications provides
access to data anywhere and anytime.
With support for small-packet protocols,
HMIs can talk directly to IIoT devices and
IT-based protocols enable them to communicate with higher-level systems. OPC
UA allows communication between an
HMI and the other devices, requiring more
bandwidth.
These new HMIs are becoming more data
aware, more application diverse, and more
process integrated and adaptive. They are
poised to play an even more capable role
within automation system architectures,
with flexible integration with the cloud,
and future machine learning and artificial
intelligence becoming a common theme. EP
Marcia Gadbois is the President and General
Manager, ADISRA, Austin, TX (adisra.com).
An entrepreneur who grew a start-up from
inception to a successful liquidity event, prior
to joining ADISRA, Gadbois was a contributing author to the book, Client/Server Programming With RPC and DCE.
EFFICIENTPLANTMAG.COM |

1/30/20 9:50 AM

http://www.adisra.com http://www.EFFICIENTPLANTMAG.COM

Efficient Plant February 2020

Table of Contents for the Digital Edition of Efficient Plant February 2020