4 6 · d e t e k t o r i n t e r n a t i o n a l
The interview
Intelligent edge-based video surveillance allows
the system to become faster, more scalable and
more customer friendly, according to I-Pro Chief
Operating Officer, Gerard Figols.
In the future we will regard cameras as AI
agents. They will have power to be able to see,
listen, and to understand what is going on. It is
important they are continually trained, enhancing
their intelligence, which means their algorithms
are becoming smarter, he says.
"Cameras will become AI agents"
only the metadata the key
information rather than the
full video stream. This allows the
system and the operator to focus
immediately on what matters and
to make faster, more informed
decisions."
What other benefits does at the
edge bring?
"It allows the system to become
faster, more scalable and more
customer friendly including
better privacy and ethical AI use.
We are proud to be the first in
the physical security industry to
achieve ISO/IEC 42001 certifica-
tion, ensuring our technology is
ethically governed by design. Edge
computing also increases customer
flexibility, letting them scale from
one to thousands of devices
without needing massive server
capacity."
How are AI and machine learn-
ing being integrated into your
edge surveillance devices?
"We integrate deep learning
AI entirely into the camera using
a powerful built-in AI processor.
This means that the camera can
do more than just record video. It
can understand what it sees. For
example, I-Pro cameras can detect
and classify up to 98 different
attributes the highest in the
industry. Users can also custom-
ise the camera to their needs,
meaning that they can train the
camera to detect specific objects
or behaviours which are important
for them."
How does that work?
"The great thing about this
training is that it can be done eas-
ily: thanks to our unique on-site
learning feature, it can be done
by the customer, and it takes just
a few minutes. The camera will
remember what it has learned, and
it will become even better over
time. The algorithm is going to be
renovated continuously to make
sure that the accuracy will keep
improving and improving. So, in
short, what we are doing is turning
the camera into a smart learning
device. It understands and helps
the operators to make faster and
better decisions."
In 2019, I-Pro became independent from Panasonic.
"Edge computing not only increases speed
and scalability, but also reduces energy
consumption and the need for extensive
infrastructure," says Gerard Figols.
Gerard Figols, I-Pro:
Gerard Figols joined Panasonic in
2007 and was leading curve-out to
establish I-Pro in the EMEA Re-
gion in 2021. His role has evolved
from EMEA to global responsibili-
ties being now Chief Operating
Officer.
Can you explain what intelligent
edge-based video surveillance
means?
"Intelligent edge-based system
surveillance means that the camera
itself can process and analyse the
video data in real time. And that
is important. So basically, what it
does is analysing the data where
the data is created itself. And this
is a big difference from traditional
systems where video is sent to a
server or a cloud platform to be
processed. Edge computing not
only boosts speed and scalability
but also reduces energy consump-
tion and infrastructure require-
ments. Importantly, it strength-
ens privacy by minimizing the
transmission of personal data to
centralised systems, making GDPR
compliance easier."
What is the biggest difference
from traditional systems?
"So in those traditional setups,
you often need a lot of network
bandwidth, and you need a very
powerful backend system. Some-
times the response can be delayed
because there is latency between
the data being sent to the data be-
ing analysed. With the edge-based
systems like I-Pro is developing, all
the AI applications and processes
run inside the camera, enabled by a
strong AI processor. This allows for
real-time alerts and actions faster,
more secure, and more efficient.
We believe it is a shift from passive
to proactive devices."
Does on-device processing of raw
data enhance the precision of
analytics?
"Yes, when you analyse the
data at the edge, it is the raw
data. It allows you to be more
accurate, because the image is not
compressed. When you compress
data, you always lose something.
It would be too heavy and costly
to transfer all this data to a server
and process all the data for many
different devices at the same time.
What is really important is that
by analysing the data directly at
the edge, our cameras can send