CS USM TRIZ Slide-proposed

MyTRIZ Competition 2012Proposal Team:

CS Aisyah Ismail

Azleena Mohd KassimMohd Adib Omar (Group Leader)

Tan Choo JunWan Mohd Nazmee Wan Zainon

Contents

• Part 1• Question (a)• Impact: The country will have sustainable

skilled and qualified ICT workers needed for the future.

• Part 2• Impact: The wood materials are transported

without damaging the pipe.

• Part 3• Impact: Enhancing the FAM Learning

Mechanism using Evolutionary-based TRIZ S-Curve Model

Part 1: Problem Analysis

• Shortage of skilled ICT graduates

• Why? Because current Institute of Higher Learning (IPT/IPTA) is unable to produce sufficient "skilled" graduates

• Why? Because good students tend to choose other fields, e.g. Medicine, Dentistry & Pharmacy etc.

• Why? Students believe that career in ICT does not provide good pay and job satisfaction

• Why? Misperceptions that ICT field does not provide better pay, low jobs satisfaction and lacks of jobs opportunities.

TRIZ Process, TRIZ Tools & Potential Solutions

Problem Insufficient Awareness and Wrong Perception on ICT Career

Insufficient ICT Teachers

Computing Major is not as attractive as Medicine, Engineering,

Accounting and Pharmacy among best students

Minimal entry requirement

No cap on the student limit per intake

Many unemployed due to skill

mismatch

Stages of Educational Process

Primary and Secondary

Education

input Pre-University

Matriculation

Diploma

Form Six (STPM)

input University input Industry

Government

Further Study

Entrepreneurship

TRIZ Process and Tools Physical Contradiction by Separation in Time Physical Contradiction by Separation in Time Physical Contradiction by Separation in Space and

Physical Contradiction by Separation in Time

Physical Contradiction by Separation in System Level - Supersystem

Inventive Principles 9. Preliminary anti-action 10. Preliminary action 13. ’The other way round’

20. Continuity of Useful Action

19. Periodic Action

23. Feedback

Proposed Solutions 9. Cultivate Problem Solving Culture

9. National Computing Competition

9. Awareness Program: ICT camp

9. Increase ICT Teachers

10. Cultivate Algorithmic Thinking

10. Standardized and up-to-date ICT Curriculum

13. Stringent entry requirements

13. Limit intake to ensure quality

20. Industrial Training & Internships

19. Periodic Retraining the

unemployed for certification

23. Formation of ICT Professional body to monitor the best practices, curriculum and prestige

Expected Outcome inculcate the right perception to students about Careers in ICT Students who are interested in ICT are prepared for university

education

Committed students

Quality ICT Program Employable graduates by the

industry

Part 2: Problem Statement

• The 1-meter pipe leaks due to the rubbing and knocking of wood materials transported inside the pipe using water as a medium of transport.

Pipe (with water) to transport wood material

example of wood material

TRIZ process used

Function Analysis:

Pipe contains water and wood based materials

Water holds the wood based materials inside the pipe

Water flows with wood based materials through the pipe

Identify the type of contradictions:

Administrative contradictions: We want to allow the wood based materials to traverse through the pipe without damaging the pipe.

Engineering contradictions: wood based material surface contact damages the pipe

Physical Contradiction Solution Strategies: Separation in Space

Inventive Principle:

30. Flexible Shells / Thin Films

TRIZ tools used

• Contradiction Matrix cannot be used since we are dealing single parameter, in this case, the contact surface of wood based materials.

• Instead, Physical Contradiction Solution Strategies: Separation in Space is used. It leads to the Inventive Principle: 30. Flexible Shells / Thin Films

Potential Solutions

• Cover the wood based materials with bubble wrap and then transport them through the pipe.

Part 3:

Enhancing the FAM Learning Mechanism

using Evolutionary-based TRIZ S-Curve Model

The Scope ofTRIZ Hybridization Model

1.A combination with a Pareto-based algorithm (MmGA) to undertaking the efficiency of learning process in FAM

2.An optimization of the number of nodes in FAM network without prior configuration towards Pareto Front solutions

3.An implementation of proposed model using USM Extract with Mobile Desktop Grid (MDG) in the form of web application and API level

Do you see the their differential?

FAM

Fuzzy Set ARTMAP

ART-1 ART-2

Supervised

Learning

Model

Unsupervised

Learning

Model

The Fuzzy ARTMAP (FAM)

FAM (same as ART and ARTMAP) consists of

Training Phase Prediction Phaseand

Note:They also called as Training Model and Prediction Model

The FAM Architecture

Nodes are created when sensation and expectation does not exceed the vigilance value

Training weights (knowledge) are kept for Prediction Phase activities

Adapted from: G. A. Carpenter, S. Grossberg, N. Markuzon, J. H. Reynolds, and D. B. Rosen, “Fuzzy artmap: A neural network architecture for incremental su-pervised learning of

analog multidimensional maps,” IEEE Transactions on Neural Networks, vol. 3, pp. 698–713, Sept. 1992.

The Optimization Dilemma and FAM

• The accumulated knowledge of FAM in Optimization Dilemma

The TRIZ-basedEvolutionary S-Curve

• The metrics of TRIZ’s Evolutionary S-Curve

The FAM and TRIZ-based Evolutionary S- Curve

• Metrics of determining the FAM's knowledge laying alone the TRIZ-based Evolutionary S-Curve

1.Performance

2.Number of Inventions

3.Level of Inventions

4.Profitability

5.Cost Reduction Related Inventions

The USM Extract

The form of USM Extract Implementation with MDG

• The flow of

• Web Application Instance

The flow of

API Implementation

The Case Study of Our Model

Prediction Phase result for Diabetes patient

Thank you