Once tasks (units in project) are identified during WBS process, individual tasks should have dependency or order. In scheduling process, program managers have to order the tasks, by identifying some can be parallelized, some should be sequential. Most importantly, tasks have to be identified their resource and duration. This normally needs the inputs from functional managers and their team members. Experience level and numbers of resource (man power) usually determine the task duration, however, it is always important to consider appropriate numbers and level, allocation (assignment), work environment such as equipment availability. It is highly recommended that project managers communicate with functional managers or lead-engineers.
1. Baseline development (One engineer – 2 weeks)
- New make target development
- Linked to new kernel library
- Minimum frame work ported
3. Module base porting (5 engineers – 2 weeks)
- Module A porting and unit testing
- Same for B, C, D, E
5. Testing result review and bug fix (? , ?)
Task 5 includes unknown factor and needs risk management. To schedule, use historical data and put conservative number, monitor closely.
(1 hour)
How to increase the Accuracy in Estimation? The followings are keys to increase accuracy in project duration estimation:
- After WBS, have the experts to estimate tasks durations.
- Refer to the previous projects which is similar. (= Analogous Estimating)
- Allow for a reserve in the estimate to provide some level of padding in case of unknowns. (Conservative Estimation)
- Create some major milestones in the project to get the idea of overall progress, make easy communication among departments.
(2-3 hours)
"Quality Triangle"
This shows the trade-offs inherent in any project.3)
/ \
C -- Q
Case2: Quality->Fix { cost increase; time decrease } or { cost decrease; time increase }
Case3: Cost->Fix { Time increase; Quality increase} or { Time decrease; Quality decrease }
Case4: Cost and Time->Fix { Quality increase by reducing features } or { Quality decreasing by adding more features }
Balancing (ideas from textbook):
- Project level
Outsourcing ( Has Negative/Positive effect)
- Business case level
- Enterprise level
Oversea design centers
-> use the equipment during day and night (No down time), experts with cheap labor
Prioritize the projects, focusing on one project
Improve working environment to be productive, introduce tools or system.
1) Refaat H. Abd El Razek, Ahmed M. Diab, Sherif M. Hafez, Remon F. Aziz “Time-Cost-Quality Trade-off Software by usingSimplified Genetic Algorithm for Typicalrepetitive Construction Projects” (2010) World Academy of Science, Engineering and Technology 61
2) http://en.wikipedia.org/wiki/List_of_project_management_software
3) http://www.projectsmart.co.uk/project-management-scope-triangle.html