Image Image Image Image Image Image Image Image Image Image
Scroll to top


Computer-Aided Innovation Conference in Harbin, China: Day 1, Afternoon Session

| On 20, Aug 2009

Ellen Domb

Report on the Computer-Aided Innovation conference in Harbin, China.  Day 1, afternoon session.

Two parallel sessions for the afternoon gave a wide variety of speakers an opportunity to tell the conference about a wide variety of research in both TRIZ methods and CAI applications, in 10 minutes each.  A few examples:

The group from Hebei University  Mechanical Engineering presented a method for fusing plastic pipe based on TRIZ.   They pointed out that the problem is not a simple contradiction:   they used the 40 principles, the separation principles, and 5 of the standard solutions to develop a real breakthrough solution

Denis Cavallucci presented a system for monitoring the impact of solution concepts within a given problem.   He illustrated this with a case study done by INSA with Arcelor Mittal, dealing with steel roll processing.  He showed how graph theory can be used to analyze the function model and the cause-effect diagram to identify the problem that has the highest priority for solution, and will remove the greatest part of the over-all problem.   90 contradictions came from 30 problems, and were resolved using 4 techniques.    The system is self-improving, since the solutions are categorized and matched to the graphs, and the confidence of the researchers increases with the improvement of the system.   This improves the willingness of the group to use innovative methods.     

Prof. Derek Tate from Texas Tech University presented work by Christopher Adams and himself on Computer-Aided TRIZ Ideality and Level of Invention Estimation using Natural Language Processing and Machine Learning.   This is part of a project to give feedback to innovators for improving the probability of success of an invention.    I predict a lot of interest in this work, and a lot of future development.

TRIZ-based hybridization of topologically optimized density distributions was presented by Gaetano Cascini, on behalf of co-authors A. Cardillo, F. Saverio Frillici, and F. Rotini.  The overall goal is to define a paradigm for CAD systems that overcomes the rigidity of conceptual design that locks out inventive concepts early in the design process.   They obtained in one step the same result that took 600 generations using a genetic algorithm, for examples of wheel design and plate design, and a number of other simple structures.  He proposes applying the density distribution method in combination with genetic algorithms to explore a more complete solution space.  

Lihui Ma presented the paper from a group at Hebei and Tianjin: UXDs-driven transferring method from TRIZ solution to Domain Solution.   This is a problem for many people learning TRIZ.   (UXD’s are unexpected discoveries.)   He presented an example that used their advanced algorithm to create a domain solution for the problem of dispensing traditional Chinese medical products, which are very viscous, through a small dropper, which gets blocked by the product.   The early results are very encouraging.

Humberto Aguillo presented his paper with Noel Leon on Predicting Innovation Acceptance by Simulation in Virtual Environments.   Virtual Selection is an emerging concept that eliminates poorly-performing variants early in the development process. Toyota’s success with the Scion, tested and refined in the virtual marketplace, was shown as an example, and explained with the Kano model.   A current example is the Future Green Chatham homes now being tested in the Etopia community on Second Life. Humberto presented a plan for future research that will combine the results from the customer experiences with the strategies for genetic variant development.

One sign of a good conference is that the coffee breaks become debates…we reconvened for the second session.  Reminder:  since I could only be in one room at a time, these are just examples of the papers.   Please see the full program for the many excellent papers that I missed.  

Fuying Zhang reviewed the concept of Ideality and the IFR, and the challenge of designing to achieve ideality.   Her research proposes a 4 stage design process familiar to TRIZ practioners that starts with formulating the IFR, analyzing the problem, resolving the problem, and then improving the solution to increase ideality. Her example of a reciprocating seal was very clear, and showed how the focus on ideality prevents the designer from using complexity to solve the problem. 

Weidong Yang presented the application of TRIZ in patternless casting manufacturing.  Patternless casting refers to systems in which a device is deposited or grown, rather than cast.  There is a pattern, but not a mold – most rapid prototyping systems and 3D printing systems qualify as patternless, in Yang’s terminology.   The practical problem addressed was getting the resin content right for the product.  

The final presentation of session 3 was Yan Yang’s study on the application of extended matrices based on TRIZ in constructing a collaborative model of an enterprise network. Her group is at the University of Electronic Science and Technology of China.   They applied the concept of contradiction to the concepts of trust, control, relationship and interaction which their research identified as the essential parameters of enterprise network function.  

The afternoon concluded with the brief business meeting of IFIP working group 5.4.   The new chairman will be Gaetano Cascini, who is also current president of the European TRIZ Association, and a well-know author for The TRIZ Journal.  Congratulations, Gaetano!    And many thanks to Noel Leon for his chairmanship.    After dinner, Prof. Tan hosted a tour of Harbin for some of the speakers who had arrived too late for the pre-conference tour.

[IMG height=128 alt=”” src=”” width=448 border=0]

[IMG height=342 alt=”” src=”” width=336 border=0]

[IMG height=480 alt=”” src=”” width=640 border=0]