Asia-Pacific Broadcasting (APB) @ConnecTechAsia Show News - Day 1 | Page 32

32 TUESDAY
@ConnecTechAsia2018
26 June 2018
www.apb-news.com
Leverage AI & foster young talent
BY MARTIN COLEMAN
Ask anyone from any industry vertical
what the current game changer is,
and I am fairly certain that artificial
intelligence (AI) and machine learn-
ing will figure in their response. It is
hardly surprising when you consider
the pure breadth of potential it af-
fords across multiple applications.
Satellite is no exception, yet we have
not yet seen many examples of AI
being used to improve processes
and efficiency, and I think that really
needs to change.
An ageing industry
Go to any satellite industry event and
two things stand out: first, there is a
lack of female presence, a topic for
another time, perhaps; secondly, we
have a distinct lack of young talent.
This is going to be a problem when
you consider the wealth of accu-
mulated experience built up from
those senior engineers. As they get
to retirement age, a great deal of
knowledge will be lost.
The satellite operators already
put a great deal of effort into train-
ing, and I know some are looking
to attract younger talent. However,
that is a very lengthy process: it will
take nearly three years to go from
a middle-tier engineer to someone
who is top of their game, and that
is expensive.
A digital assistant
AI can help here. There are two very
distinct methods of AI: knowledge-
able engineering, which is about
teaching a machine some of the
knowledge accumulated by people;
and machine learning, where we
teach a machine the facts and by
feeding it information and the use of
probability and uncertainty process-
es. A combination of these methods
will have a massive impact on improv-
ing efficiency and reducing errors,
including interference of course, for
satellite operators.
If we feed that machine informa-
tion about everyday satellite process-
es, service details, scheduling, issues,
and so on, we can begin to automate
everyday processes, whether that is
procedural, error resolution or risk
management. The more information
fed into the system, the more accu-
rate it can be and more processes can
be automated. I view this like having
a digital assistant for the satellite
industry. For me, with my focus very
much on interference resolution, I
see endless possibilities for a machine
that can automate the steps needed
to resolve and predict interference
before it occurs, based on re-occur-
ring themes and patterns. However,
the beauty of it is that it could be
taught a whole spectrum of things
relating to the day-to-day tasks of
satellite operations engineers.
What it never can be is a replace-
ment for people, as you still need that
human element to verify that right
choices are being made and inter-
vene, and continue the feeding and
training of that machine. So, along-
side AI, we should, of course, still be
looking at education, and what we do
to attract new talent to the industry.
An industry effort
All satellite operators could begin by
implementing their own AI initiatives,
and some may well have begun that
process. However, when you consider
that many of the processes will be
standard across all satellite operators,
it would be much more efficient to
have a coordinated effort to develop
an industry-wide digital assistant that
can then be adapted by each oper-
ator as required. I would love to see
an industry effort around this.
❝ Alongside AI, we should,
of course, still be looking at
education and what we do
to attract new talent to the
industry. ❞
— Martin Coleman,
Executive Director,
Satellite Interference Reduction
Group and an APB Panellist