Presentation is loading. Please wait.

Presentation is loading. Please wait.

1 1 Case Example of Applying BridgePoint to OO Development Training Shohei KUKI Ricoh Company, LTD.

Similar presentations


Presentation on theme: "1 1 Case Example of Applying BridgePoint to OO Development Training Shohei KUKI Ricoh Company, LTD."— Presentation transcript:

1 1 1 Case Example of Applying BridgePoint to OO Development Training Shohei KUKI Ricoh Company, LTD.

2 2 2 KUKI, Shohei Started working as an embedded system engineer at Ricoh in 1999. First project was UI (User Interface) common component development, where he gained experience and skills on object-oriented software modeling. Has been training new engineers to improve their abilities and skills in object-oriented software modeling, as well as creating a system of skill development in object-oriented software modeling since 2003.

3 3 3 Contents Background Introductory OO Development Training Road to the effective training –Use BridgePoint –Lecture the concept of domain –Iteration evaluation review Conclusion Case Example of Applying BridgePoint to OO Development Training

4 4 4 Background History Practical OO Development Training Initial skill level of trainees

5 5 5 History 1996 20012002 Trying to adopt MDD again 2003 Model based OO development(the differential development) Rose was chosen as the major tool. Practical OO development training for newly hired employees Modeling technique was learned from SMM at the beginning. Bridge Point was one of the proposed tools. Bridge Point was not chosen as the major tool. but, still automatic code generation was promising. developed a part of printer engine controller by Bridge Point with a consultant. MDD was not adopted. OMG began advocating MDA. 20002004 Then,proposed MDD to a production side.

6 6 6 Practical OO Development Training Purpose –Developing software modeling engineers –Making constitution of building the quality into software at upper process –Making a set of software as outputs Trainees –Newly hired employees who assigned as software engineers Period –About 1 year

7 7 7 Initial skill level of trainees Academic background –University or master course graduates Programming experience –Some have development experience, some have no knowledge of programming language example in 2004: Experience of OO development –Nobody has knowledgedevelopment C 97 % 59 % C++ 59 % 25 % Java 59 % 16 % Other language 22 % 9%9%

8 8 8 Introductory OO Development Training Positioning and purpose Spec of Training Spec of Subject Development plan Training style and System

9 9 9 Positioning and purpose Positioning –Initial training of “practical OO development training” purpose –Aware of the fun in developing embedded software –Aware of the base of OO thinking

10 10 Spec of Training Subject Pathfinder ※ Hardware LEGO MindStorms OS legOS 0.2.5 Trainees Newly hired employees who assigned as software engineers initial skill of trainees Exp. of software development: about 70% Exp. of OO development:0% Period About 10 days ※ This subject is the same as the one used for UML robot contest. (requested spec is simplified) URL of UML robot contest: http://www.otij.org/umlforum2004/robocon/ (Japanese Version Only)

11 11 Spec of Subject Pathfinder –Self-steering vehicle –running along the line Proceed Stop Pause for 2 seconds Detect line Detect out-of-track –Maintenance Correct threshold level of sensors Check motors

12 12 Development plan Achieving functions incrementally Basic func.Proceed, stop and detect line Additional func. 1 Detect out-of-track (in 2002 and 2003) Pause for 2 seconds (in 2004) Additional func. 2 Check motors (in 2002 and 2003) Correct threshold level of sensors (in 2004)

13 13 Training style and system Training style –Lecture + exercise(group work) Promote understanding by many exercises Training system –lecturer Give lectures –advisers Support reviews 2 advisers for about 3 groups –trainees Form a group with 4 to 6 members

14 14 Road to the effective training

15 15 We did various kaizen rate of achievement of total groups year functions 200220032004 Basic func.100% Additional func. 180%100% Additional func. 20%80%100% UP! Not good progress Little time for analysis Not best progress Kaizen1: use BridgePoint Kaizen2: Lecture the concept of domain Kaizen3: Iteration evaluation review

16 16 Result in 2002 No group could achieve additional func. 2 rate of achievement of total groups year functions 2002 Basic func.100% Additional func. 180% Additional func. 20% No group could achieve additional function2 despite my efforts in preparing this function!

17 17 Problem in 2002 problem –Not good progress Many trainees deadlocked at design and implementation –Some trainees didn’t have experience of programming –Little time for analysis cause –Mismatching between training level and trainee’s skill level Especially, design and implementation were difficult for the trainees We used Rose as development tool, with which implementation is necessary

18 18 Kaizen1 in 2003: using BridgePoint Changed development tool –Switched from Rose to BridgePoint With BridgePoint, design and implementation is not necessary

19 19 Result in 2003 Speed up of development rate of achievement of total groups Had a lot of time for analysis year functions 20022003 Basic func.100% Additional func. 180%100% Additional func. 20%80% UP! But some groups could not achieve additional function2

20 20 Problem in 2003 problem –Not best progress Couldn’t separate semantics –pathfinder and device Model quality varied depending on each group cause –Didn’t lecture the concept of domain Concept of domain and bridge was not understood well –Lack of quality management Couldn’t take hold of each group

21 21 Example : mixed semantics

22 22 Kaizen2 in 2004: lecture the concept of domain Added the concept of domain lecture to program –Increased training days, lecture and exercise

23 23 Example: separated semantics

24 24 Problem in 2003 problem –Not best progress Couldn’t separate semantics –pathfinder and device Model quality varied depending on each group cause –Didn’t lecture the concept of domain Concept of domain and bridge was not understood well –Lack of quality management Couldn’t take hold of each group

25 25 Kaizen3 in 2004: Iteration evaluation review Iteration evaluation review –Review for outputs in each iteration –We can take hold of each group’s status and quality of outputs. –Late group can increase the speed by doing some review intensively

26 26 Result in 2004 All groups achieved all functions! –Could shift the semantics –Developed the models in excess of a certain level rate of achievement of total groups year functions 200220032004 Basic func.100% Additional func. 180%100% Additional func. 20%80%100%

27 27 Comments from trainee Satisfaction level No data Comment sample Very satisfactoryI learned how we should do the group work Very satisfactoryPathfinder was a fun as a subject Very satisfactoryNow, I can understand the flow of OO development

28 28 We accomplished the purpose Aware of the fun in developing embedded software –All groups achieved all functions Aware of the base of OO thinking –Focused on analysis –Could shift the semantics –Developed the models in excess of a certain level

29 29 Conclusion

30 30 What we learned about education Take care of trainee’s skill level ! –Newly hired employees may not have implementation skill Focus OO thinking ! –Take a lot of time to analyze –Beginners cannot separate semantics Make trainees be fun to develop software ! –Rapid development BridgePoint is an excellent tool for education and future development

31 31 Thank you for listening! Case Example of Applying BridgePoint to OO Development Training Shohei KUKI Ricoh Company, LTD.


Download ppt "1 1 Case Example of Applying BridgePoint to OO Development Training Shohei KUKI Ricoh Company, LTD."

Similar presentations


Ads by Google