IEEE Computational Intelligence Magazine - August 2022 - 6

... Create a forum within CIS conferences, in which
Industry can participate without having to submit
scholarly papers. Established Industry Day within the IEEE
WCCI 2022 Program. Organizing the 2023 IEEE Conference
on AI as an industry-centered conference.
value at each step of the data journey
and be ahead of the competition.
❏ AI-enabled Cybersecurity and Privacy:
This panel will share real-world challenges,
novel applications, and the
most recent advances in security AI,
and privacy from theoretical and
empirical perspectives. It will address
questions such as: How can we make a
system robust to novel or potentially
adversarial inputs? How can machinelearning
systems detect and adapt to
changes in the environment over
time? When can we trust that a system
that has performed well in the past
will continue to do so in the future?
❏ Industrial AI: This panel will cover
industrial applications of AI, ranging
from production to manufacturing,
quality control, operations, and
maintenance. We will show illustrative
examples from large and small
companies. We will extend the discussion
beyond the underlying AI
technology to cover Business and
Technology Challenges. We will
explore Business Models to monetize
AI-improved processes or AI-enabled
services or products; KPIs for AIenabled
process improvement; Productization
of AI-enabled services
and products; AI Technology development
and deployment; Technology
sustainability, such as Model maintenance,
etc.
❏ Technology and Regulatory Challenges
for AI in Europe: We will share the latest
policy document on AI of the
IEEE European Public Policy Committee
ICT Working Group and discuss
the technology and regulatory
challenges for AI in Europe, hearing
from experts from Academia, Industry,
and EU Policymakers.
❏ Applied Deep Learning in Satellite
Images using Small Supervised Data:
Deep Learning technologies need to
have annotated data to train their
models. This issue is particularly
severe in operational scenarios where
it is not possible to annotate data in a
reasonable time frame or in cases in
which we do not have enough available
data related to rare scenarios.
This panel will deal with methodologies
for training neural networks
with little data, using Self-Supervised
Learning techniques.
❏ AI Techniques for Natural Language Processing
(NLP): Questioning Evaluation
and Resource consumption for
Deep Language Models in realworld
scenarios. Large Language
Models (LMs), currently reaching a
trillion parameters, need reliable,
complete, and representative benchmarks.
This task usually requires significant
effort in text annotation,
intensive in both time and required
knowledge, except for simple cases.
These benchmarks are far from real
industrial NLP use cases. Moreover,
some LMs provide limited benefits
to NLP tasks, where sometimes
cheap, known, ML techniques are as
effective with a tiny fraction of cost/
time/resources/carbon footprint.
This panel will focus on how to
understand in advance when such an
effort can bring significant benefits.
❏ AI for Earth Observation: AI for Earth
Observation (AI4EO) focuses on
harnessing the power of AI with the
vast amount of EO data now available.
While today the new boost of
AI4EO is mainly related to Computer
Vision applied to high-resolution
satellite imagery, there are many
other areas for Earth Science, prediction
and big data analytics that could
benefit from AI for the deployment
of practical applications. The panel
will discuss some of the main challenges
in AI for Earth Observation
of interest for research and business
communities, i.e., scalable big data
analytics, trustworthy and explainable
AI, physics-aware AI, self-learning
AI, AI-based EO data fusion and
prediction for Digital Twin Earth
(DTE). Here are examples of issues
that will be discussed: How to augment
EO capability with AI? How
to make the AI decision-making
more transparent? How to integrate
" first principles " and " domain
knowledge " into the AI statistical
approach? How to develop unsupervised
learning for EO data without
labels? How to leverage the combination
of EO, models and AI techniques
to support high-resolution
prediction and informed decision
making with DTE?
The interested reader can find additional
information on this event at
https://wcci2022.org/industry-day/.
Our second effort in industrycentered
conferences is the organization
of the 2023 IEEE Conference on
AI (IEEE CAI) - https://cai.ieee.org/
2023/. This conference, co-sponsored by
the Computational Intelligence, Computer,
Signal Processing, and Systems
Man and Cybernetics Societies, will take
place on June 7-8, 2023 in Santa Clara,
CA, USA. IEEE CAI 2023 will be the
first full-fledged conference following
this new format. Different Industry segments
have different needs, requirements,
regulatory constraints, and levels of technology
adoption. To reflect this heterogeneity,
we will structure IEEE CAI
along six verticals, covering AI in Energy,
Healthcare/Life Science, Transportation/Aviation,
Earth System Decision Support, Industrial
AI, and Social Implications of AI/Privacy.
As we get closer to that date, we will
provide additional information in our
next columns.
6 IEEE COMPUTATIONAL INTELLIGENCE MAGAZINE | AUGUST 2022
https://www.wcci2022.org/industry-day/ https://cai.ieee.org/2023/ https://cai.ieee.org/2023/
IEEE Computational Intelligence Magazine - August 2022

Table of Contents for the Digital Edition of IEEE Computational Intelligence Magazine - August 2022
