Social Cyber System and Machine Learning

in Era Society 5.0

Dr. Filbert H. Juwono
PhD, MEng, BEng, SMIEEE, FHEA

Curtin Malaysia University is owned and managed by Curtin (Malaysia) Sdn. Bhd. (464213-M)
Ministry of Education Registration No. KPT/JPT/DFT/US/Y02 DULN003(Q)
CRICOS Provider Code 00301J.
Curtin’s largest international campus and global hub in Asia
 FACULTIES
Faculty of Engineering and Science

Faculty of Business

Faculty of Humanities
Industrial and Society Revolutions

Source: ANZ
Society 5.0
What is the difference?
What is Society 5.0?
• Merging between cyberspace and physical space
• Vision of Society 5.0 reframes two relationships:
• Relationship between technology and society
• Technology-mediated relationship between individuals and society
• Basic scheme:
• Data are collected from “real world”
• Processed by computers
• Results being applied to the real world
• Simple example: air-conditioning which keeps a room at the desired
temperature
It is more than that …
• Society 5.0 will have systems that operate throughout society in an
integrated fashion
• We require comfort in all aspects of life, including in energy,
transport, medical care, shopping, education, work, and leisure
• Data are processed by sophisticated IT systems such as Artificial
Intelligence
• Society 5.0 will directly shape our actions and behavior
Society 5.0 will feature an iterative cycle in which data are gathered, analyzed, and then converted into
meaningful information, which is then applied in the real world; moreover, this cycle operates at a
society-wide level
Merging Cyber Space and Physical Space
• Cyber space: a digital space in which real-world data are collected and
analyzed to derive solutions
• The infrastructure of this space is the vast array of computer
networks
• Physical space: real world, from which raw data are collected and into
which solutions are applied
• Merging cyberspace with the physical space: a cycle in which data
smoothly flow from the physical space (real world) into cyberspace
and then flow back from cyberspace into the physical space (real
world) in the form of meaningful information
Example Scenario
Toward a People-Centric Society
• Society 5.0 will become people-centric society
• Japan’s Government strategy:
balance economic advancement with the resolution of social problems … to ensure
that all citizens can lead high-quality lives full of comfort and vitality
• Dilemma:

Quality of life
Environmental issues

• Social problems?
Convergence and the Benefit to the Society
• Cyberspace must have a structure mirroring that of the real world
• Example: air conditioner
• the cyber model must have a real-world mirroring structure necessary for air-
conditioning the room
• system must model the physical characteristics of the room to understand
how the room will change if the airflow is increased or decreased
How?
• How to get the “data” of the room?
• IoT which utilizes sensors
• How to analyze the system to get the decision?
• Artificial intelligence (AI)
• AI notes the mistake and readjusts the model accordingly
• How does Society 5.0 differ from Cyber-Physical Systems (CPS)?
• Society 5.0 does not involve separate, isolated systems
• Society 5.0 is cyber-physical convergence at the level of society as a whole
Interconnection
• Each system will become interconnected in cyberspace
• How different services interconnect?
• How will this insight benefit society?

Energy Transport Water

Cyberspace

Public
Healthcare Distribution
security
Example: Traffic Congestion
• What may cause traffic congestion?
• Poor public security
• Poor water infrastructure (flooding when raining)
• Illegal parking
• The entire city can be modeled in the cyberspace to analyze the root
of causes
• We can identify the lowest cost solution to mitigate the congestion
• In other words, AI can spot connections that a human would
overlook.
• With such AI, we will learn how different services in a given area
interact in the short term, and how a given service would shape other
services over a longer time span
Sharing Knowledge
• By modeling a city or society in cyberspace and linking it with other
cities or societies, it will be possible to extrapolate existing knowledge
• If we have analyzed some data pertaining to a given city using a
certain method. This method may be applicable to another city
Other point of view of Society 5.0
• Knowledge-intensive society
• Data-driven society
Knowledge-Intensive Society
• Society 5.0 identifies
three elements that
drive social
innovation:
• Data
• Information
• Knowledge
Definitions
Data is tangible and intangible phenomena in the physical space (real world) that are represented as numerical
values, states, names, or binary figures (0 or 1) telling us whether a thing is present or absent

Information is data that has been rendered meaningful by selecting and processing it for a particular purpose
or as part of a course of action

Knowledge is what enables you to make a decision

What is a Knowledge-Intensive Society?

In Society 5.0, the process will be driven without human intervention; of the three
elements, humans will only gain greater opportunities to access AI-derived
knowledge, the final output of the conversion process.
Data-driven Society
• Data exist at a more basic level
• Data-driven society is a society where data (gathered by IoT
networks) are converted into information and knowledge, which then
“drive” the real world
• In the future, AI will convert data into knowledge (general empirical
laws) through an automated process, and then use this knowledge to
automatically control devices.
Healthcare in Society 5.0
Some Scenarios
• Personal data store managed through biometric authentication
• Dietary recommendations
• Remote monitoring services
• Design of personalized healthcare plans
• Health improvement incentives
• Offshore healthcare services
• Physical/mental support and psychiatric care
Artificial Intelligence
• Artificial Intelligence (AI) is the development
of computer systems that are able to
perform tasks that would require human
intelligence.
• Visual perception
• Speech recognition
• Decision-making
• Everyday examples: chatbots, human vs
computer games
• Weak vs strong AI vs super AI
 Machine Learning
• A subset of AI
Artificial • Algorithms for learning the data
Intelligence

Machine
Learning

Deep Deep Learning

Learning • A subset of Machine Learning
• Neural networks for learning the data
 Machine Learning

Supervised Learning Unsupervised Learning Inference

Classification Regression Deep Learning Clustering

Linear
Naïve Bayes ANNs K-means
Regression

Neural Hidden
kNN Networks CNNs
Markov

Decision
SVM
Trees
Smart City, Smart Grid, and Society 5.0

Society 5.0

Smart Grid Smart City Supersmart Society

AI in Smart Grid
• Smart grid integrates two-way communication technologies and cyber
components, into power equipment to boast higher reliability,
efficiency, sustainability in an environment of increasing power
demand
• Power control mechanism: power disaggregation
• Intrusive load monitoring (ILM)
• Non-intrusive load monitoring (NILM) – more practical and cheaper
• Cyber-attack problems
 Load Categories
• Device categories based on the power consumption

Two-state appliances:
Bulb, toaster

Varying appliances:
Electrical sewing machines, light dimmers

Finite state machines:

Hair dryer, washing machines

Permanent consumers:
Telephone, smoke detectors

M. H. R. Bhatti and F. H. Juwono, “Power Disaggregation in Smart Grid Applications,” Internal Report, 2019
Load Categories
• Based on load characteristics
NILM Methodology
• Methodology:
• Data acquisition
• Feature extraction
• Appliance classification
• Device signatures
• Set of features of electrical appliances: signatures
• Signatures can be acquired when the device is in steady-state or transient
state
• Load features captured during the steady-state have negligible variation:
Power (P, Q), RMS current and voltage, harmonics, V-I curve, power factors
kNN
• Based on
majority voting
• Example:
Data
• Reference Energy Disaggregation Data (REDD)
Methodology
 Results
Split data vs accuracy
Split Data
(%)

Threshold vs accuracy
Thank you
http://gecost.curtin.edu.my