Artificial Intelligence Overview

Artificial Intelligence (AI)

• AI is the study of how to make computers do tasks that humans currently excel at.
• Rich & Knight (1991) defined AI as such.
• AI can be categorized by thought processes, (thinking like humans, thinking rationally) and actions (acting like humans, acting rationally).
• This is displayed in a 2x2 matrix.

Industry Evolution

• Historically, industries have evolved through stages, from mechanization (1st) to computer automation (3rd) and now cyber physical systems (4th).
• Each stage builds on the previous one, introducing new technologies and capabilities.
• Industry 4.0 represents the current stage.

Turing Test

• Alan Turing proposed a test to determine if a machine can exhibit intelligent behavior equivalent to, or indistinguishable from, that of a human.

• The test involves an interrogator communicating with both a human and a machine without knowing which is which.

• If the interrogator cannot reliably distinguish the machine from the human, the machine is considered to have passed the test.

• A question: Why the strange set-up of the Turing-Test? Why did Turing 'pit' a machine against a human in some kind of contest? Why not have the interrogator simply interact with a machine and judge whether or not the machine is intelligent based on those interactions?

CAPTCHA

• CAPTCHA (Completely Automated Public Turing test to tell Computers and Humans Apart) was coined in 2000 to differentiate humans and computers based on tasks that humans are better at, like recognizing distorted text.

AI Definitions

• AI encompasses a broad range of methods and systems, including: - Statistical methods such as Regression, Descriptive and inferential analytics, Bayesian networks etc. - Symbolic reasoning, rule-based systems, induction, Deduction, Fuzzy logic etc. - Bio-inspired systems, Neural networks, Genetic algorithms, Machine learning, Deep learning, etc.

The Traveling Salesman Problem (TSP)

• A classic optimization problem in AI concerning finding the shortest possible route that visits a set of cities exactly once and returns to the starting city.
• TSPs are categorized in computer complexity analysis as an NP-complete problem.

Heuristic Search Techniques

• Direct techniques (blind search) are not always effective due to high time/memory demands, and require sophisticated strategies.
• Weak techniques are effective when correctly applied to specific tasks, and often rely on domain-specific information.

A* Search

• An algorithm for finding a path with minimum cost using heuristics (estimations of cheapest path).
• The heuristic function (h(n)) estimates the cost of the cheapest path from a starting point to a goal.
• The algorithm combines the cost function of the path to a node (g(n)) with estimated cost (h(n)) to produce f(n) a function for least cost path.

Key Research Areas in AI

• Problem-solving, planning, and search are generic problem-solving methods primarily inspired by Cognitive science and robotics.
• Automated reasoning/inference uses stored information to answer and draw conclusions (data).
• Machine learning uses data to create intelligent systems able to adapt to changing circumstances.
• Natural language processing enables machine communication.
• Computer vision processes visual information.
• Robotics involves autonomous manipulation and full integration of AI capabilities.

Machine Learning

• A system with inherent ability to acquire and integrate knowledge.

Why Machine Learning?

• Necessary in situations requiring extensive expertise from humans or rapidly changing occurrences like financial models, diagnosis, and fraud detection.
• Machine learning is helpful for personalization (e.g. movie/book recommendations).

Return and Exchange Policy

• Customers return significant amounts of merchandise annually.
• Many factors contribute to returns: refunds, exchanges, unsatisfactory or faulty products etc.

Related Fields to Machine Learning

• Machine learning uses elements from diverse fields including statistics, information theory, databases, data mining, evolutionary models, neuroscience, decision theory, cognitive science, and control theory.

AI Applications in Other Industries

• AI is utilized in transportation (autopilots, self-driving cars), retail (shopping assistants), security, gaming, and financial services.

Robotics

• Advanced robotics is being developed for use in healthcare, including hospital-based robots (surgical, transportation, care) alongside home assistants and telepresence solutions.

AI Application in Healthcare

• Intelligent information gathering uses IoT devices (wearables) to collect patient information.
• Intelligent interactions enable communication with more natural systems, leading to better care.
• Diagnostic and treatment strategies are developed.
• Devices to monitor and aid medical devices.
• Robotic assists in mundane, dangerous, or complex medical procedures.
• Advanced BI/Analytics provides future predictions and actionable insights.

Evolution vs. Revolution

• Evolution is a gradual progression through different stages.
• Revolution is a rapid, radical, or fundamentally different change.

The Promise of Industry 4.0

• Evolution of technologies has resulted in the current industry 4.0 which is characterized by cyber-physical systems, computer automation, mass production using electricity and water, and mechanization.

9 Technology Pillars - Industry 4.0

• AI-driven technology pillars supporting Industry 4.0 include Autonomous Robots, Big Data Analytics, Artificial Intelligence, Supply Chains, Additive Manufacturing, Simulation / Augmented reality, Horizontal / Vertical Integration, Industrial IoT, and Cybersecurity.

Leveraging Industry 4.0 Challenges and Opportunities

• Data & Analytics, IoT platforms, mobile devices, cloud computing, and augmented reality are utilized to gain a more complete customer understanding, personalize experiences and create enhanced customer profiling. Authentication, fraud detection, and 3D printing are opportunities which address challenges faced in Industry 4.0.