Research Practices

and Atmospheres in WorldWorld-Class Research Laboratories
Asst Prof Chaodit Aswakul PhD, DIC, MIEEE Electrical Engineering Department Faculty of Engineering Chulalongkorn University

Have you done any research before?

2102790 EE Seminar, ENG CU 8Nov07

1

2102790 EE Seminar, ENG CU 8Nov07

2

RE … SEARCH
to look for (new) solutions to selected problem ... again and again to improve upon existing ones
2102790 EE Seminar, ENG CU 8Nov07 3

Searching in RESEARCH
Need of progressive searches with rewarding advancement towards successful ends Need of recursive searches with aim at improvement or refinement or corrections
2102790 EE Seminar, ENG CU 8Nov07 4

Searching in RESEARCH
Need of regressive searches to trace back to causes of failures and learn from them Need of self searches to sustain inspiration and spirit through out research period
2102790 EE Seminar, ENG CU 8Nov07 5 2102790 EE Seminar, ENG CU 8Nov07 6

Steps in ‘research’ research’

1

existing solutions? How these existing solutions work? Key logical idea = ? Where they work? (applicable range) Where they don’ don’t? (known limitations) Where they won’ won’t? (unknown limitations) Why them? (pros and cons) Why not others? (creativity) 2102790 EE Seminar.possible & reasonable Best research problem is that one you feel obsessed of solving 2102790 EE Seminar. ENG CU 8Nov07 2 . ENG CU 8Nov07 10 Second SEARCH search for data relevant to problem (literature survey) Methodology Third SEARCH search for new solutions to solve problem (proposed methodology) Analytical : mathematical analysis Experimental : implementation of idea on lablabscale/realscale/real-scale systems with evaluation by actual measurement Simulation : emulate system dynamics in computer program 11 2102790 EE Seminar. ENG CU 8Nov07 12 2102790 EE Seminar. ENG CU 8Nov07 8 Literature Review Who are main authorities in the selected problem? What are proposed. ENG CU 8Nov07 7 2102790 EE Seminar. ENG CU 8Nov07 9 2102790 EE Seminar.Problem Selection Criteria Good research problem must be First SEARCH search for problem How to select good research problem? – Worth investigating – to you. novelty – Not implausible -. to others – Not too wide – very specific objective – Not too crowded – with rooms for improvement – Not too old – originality.

ENG CU 8Nov07 14 2102790 EE Seminar. ENG CU 8Nov07 17 2102790 EE Seminar. ENG CU 8Nov07 16 Discussion Common Pitfall : After getting results. then you need more results Discussion is typically much time consuming 2102790 EE Seminar. students often think that they have finished their investigations. ENG CU 8Nov07 18 Sixth SEARCH search for conclusion (conclusive evidence) 3 . Remember : results without explanation are meaningless Need to ask why to every result If you cannot find sound reason for a result.Methodology Comparison analytical experimental Simulation Fourth SEARCH Interpretability *** ** * Cost * *** ** search for results that support new solutions (applicability range) Skill math lab programming 13 2102790 EE Seminar. ENG CU 8Nov07 Results MUST specify both pros and cons of the proposed methodology MUST be compared with existing methodologies SHOULD be based on practical ranges of parameter settings 2102790 EE Seminar. ENG CU 8Nov07 15 Fifth SEARCH search for reasons to explain obtained results (discussion) 2102790 EE Seminar.

ENG CU 8Nov07 22 Eighth SEARCH search for next problems (recommended future work) Recommendation for Future Work Good research work must open new directions for further research “Step over other giant’ giant’s shoulder” shoulder” Research community – fruitful for newcomers 2102790 EE Seminar.. ENG CU 8Nov07 23 Welcome to research world I hope you would gain much more than just knowledge 2102790 EE Seminar. ENG CU 8Nov07 19 2102790 EE Seminar. ENG CU 8Nov07 20 Publication Be prepared well ahead of submission deadline Typesetting program (LaTeX ) is (LaTeX) academically standard Conference . ENG CU 8Nov07 21 2102790 EE Seminar. Then journal 2102790 EE Seminar. ENG CU 8Nov07 24 4 .Conclusive Evidence From all obtained results … can you ‘conclude’ conclude’ anything nonnon-trivial (cannot be deduced from initial intuition before obtaining results) Seventh SEARCH search for places where you can announce research outcomes (publication) ?? 2102790 EE Seminar.

assignments and labs One comprehensive exam Plus 33-4 months to do a small research and write a thesis Study all the time (day/night/holiday) 1 year 2102790 EE Seminar. ENG CU 8Nov07 25 2102790 EE Seminar. M.a few Nobel prize winners Large research budget from both state agencies and industries Faculty staff being real experts Imperial College London (student) Top students from best of every country MultiMulti-cultural environment (Greek. Singaporean. every student is very active in asking/commenting/arguing CU classes. ENG CU 8Nov07 26 Imperial College London (MSc study) Lots of classes. ENG CU 8Nov07 27 Imperial College London (PhD study) No classes. no grade Only two exams: Thesis Proposal and Viva Time is all yours to manage Students must be independent and selfself-disciplined Up to 8 years 2102790 EE Seminar. every student is quiet At IC. time is taken seriously … deadline and schedule and plan are commitment CU time is 15 minutes late and you feel not guilty! 2102790 EE Seminar. Indian. UK and Europe Continent) Extremely hardhard-working 2102790 EE Seminar. Principles of Working Systematic study of all data Holistic view of problem Start tackling the easiest Continue in progressive steps Think out(side) out(side) of textbooks 2102790 EE Seminar. ENG CU 8Nov07 30 5 . ENG CU 8Nov07 29 H. ENG CU 8Nov07 28 Difference IC & CU IC classes.Imperial College London (institution) Number 1 in engineering of UK Science/math departments are very strong -. Chinese.

ENG CU 8Nov07 6 . ENG CU 8Nov07 H. M. ENG CU 8Nov07 หลักการทรงงานของในหลวง ระเบิดจากข้างใน ทําให้ง่าย ํ ขาดทุนคือกาไร พึงตนเอง ซื อสัตย์สุจริ ต ความเพียร 2102790 EE Seminar.burst from within Simplicity Profitable failures SelfSelf-dependence Honesty Perseverance หลักการทรงงานของในหลวง ่ นระบบ ศึกษาข้อมูลอยางเป็ มององค์รวมของปั ญหา ้ ญหาทีจุดเล็ก เริ มแกปั ทําตามลําดับขัน ่ ดตํารา ไมติ 31 2102790 EE Seminar. Principles of Working Inspiration . ENG CU 8Nov07 32 2102790 EE Seminar.M.H. King Rama IX best role model in research ่ น เราจะครองแผนดิ โดยธรรม เพือประโยชน์สุข ่ แหงมหาชนชาวสยาม 33 <what to do> <how to do> <why to do> <for whom> 34 2102790 EE Seminar.