1. A Complexity Metric for Practical Ship Design 1 Jean-David Caprace ANAST – University of Liège – Belgium PRADS – September 2010

2. Boundaries – What, How and Why? Background How to measure the ship complexity Presentation of the test case on a passenger ship Conclusion 2 Outline Of this presentation

3. 3 Boundaries – What, How and Why?

4. Ship designer problem  selection of the best design alternative Evaluation of the design alternative  many attributes (economic, technical, environmental, safety) Every design change  impact on how much producing/maintaining the ship will cost Understand the impact every time the designer make a change 4 Boundaries What? – Selection of the best design alternative

5. Complex design – Are more fragile – Leads to more surprises (always bad) – Leads to longer development schedule – Causes costly late design modification – Fosters suboptimal tradeoffs between competing goals – Makes follow-up of design more difficult 5 Boundaries What? – Selection of the best design alternative How well you handle your design comes down to how well you handle complexity!

6. To give a quantitative and objective metric for the designer – Design evaluation tool To find an alternative to the cost evaluation during the design – Expected to be faster – Easier to implement – “Real time” 6 Boundaries How? – Reduction of the product complexity

7. Design for X  Optimize total benefits – Design for production – Design for assembly – Design to cost – Design for safety – Design for environment – Design for maintenance – Design for simplicity 7 Boundaries Why? – To optimize the life cycle of the product Design for life cycle

8. complex system Very hard to find a formal definition of complex system Complexity often implies – Many parts with a lot of redundancy – Many relationships/interactions among the parts – Combination effects that are not easily predicted – A form of a hierarchy If complexity increase  LCCost increase 8 Background What is a complex system?

9. Engineers are using everyday the word “complexity” Sometimes it is easy to feel the complexity … 9 Complexity evaluation What is a complex system?

10. Engineers are using everyday the word “complexity” Sometimes not … 10 Complexity evaluation What is a complex system?

11. 11 How to measure the ship complexity

12. Complexity affects design, manufacturing, assembly operations, maintenance, dismantling, etc. Difficult to measure each factors involved in complexity assessment – Number of components – Number of connections – Number of assembly – Geometry and shape – Production processes – Density – Etc. 12 Complexity evaluation A challenge

13. Industry has already attempted to measure complexity – Using empirical measures – Problem  proliferation of possible measures number of item, production sequence and assemblies, etc. – So many metrics  how to select the most appropriate indicators ?  do you have the sufficient accuracy ?  how can you tell if the overall complexity is bring reduced if one measure falls but another rises ? 13 Complexity evaluation A challenge

14. Complexity related to … – Number of parts/connections – Complexity of each parts Different complexities – Manufacturing complexity – Assembly complexity – Process complexity – Maintenance complexity – Etc. Study limited to steel structure 14 Complexity evaluation Theoretical consideration

15. Steel structure complexity – Shape complexity (C sh ) – Assembly complexity (C as ) – Material complexity (C mt ) Complexity evaluation Different factors

16. Shape complexity (C sh ) – Ability to perform the manufacturing of individual parts of the product – Based on sphericity of the product components –  16 Complexity evaluation Different factors C sh = 0 C sh = 0.194 C sh = 0.329

17. 17 Complexity evaluation Different factors Shape complexity (C sh )  Similar to the material density  Reduce number of part and reduce the C sh of each part

18. Assembly complexity (C as ) – Ability to easily assemble the components of a product – Based on a recursive formulation similar to the Shannon Entropy – n non-isomorphic sub-trees Complexity evaluation Different factors simple complex

19. 19 Complexity evaluation Different factors

20. 20 Complexity evaluation Different factors Assembly complexity (C as )  Lifts and stairs  high assembly complexity  Use concepts of modularity and use standardization

21. Material complexity (C mt ) – Ability to use different types of material and scantling in a product  Standardization – Based on the number of different material and scantling used in the product Complexity evaluation Different factors

22. 22 Complexity evaluation Different factors Material complexity (C mt )  Material and scantling standardization is required

23. 23 Complexity evaluation Different factors Global complexity  – Weighted Sum – Minimization of the correlation coefficient – Production time vs Complexity – R² = 0.76

24. 24 Complexity evaluation Overall complexity Design manager  Define upper and lower limits

25. 25 Conclusion and recommendations

26. It’is always possible to design something so complicated that you can never get it right! This methodology provides: – An aid for the designers  compare different design alternatives – A monitoring of the sources of complexity which helps to determine the consequences of decision making – A spotting of the sources of complexity and cost which helps to reduce design effort – An objective, quantifiable, unambiguous metrics of complexity Results: – Reduction of lead time and life cycle cost 26 Conclusion and recommendations Of this presentation

27. This research have been limited to: – Ship’s structure (i.e. mainly steel parts and not outfitting) – Complexity evaluation during production (i.e. not on maintenance or dismantling) – Large passenger ships Additional researches are thus required – Outfitting components (HVAC, pipes, electrical cables, etc.) – Take into account of maintenance and dismantling stage – To test the methodology on other types of ships 27 Future work Of this presentation

28. 28 Thank you for your attention As Einstein said, everything should be as simple as possible But not simpler !