[attachment=12061]
ABSTRACT
Project planning is a business process for successfully delivering one-of-a-kind products and services under real world time and resource constraints.Developing a project plan, a crucial element of project management is a difficult task that requires significant experience and expertise. Interestingly Artificial Intelligence can be applied to develop both mixed-initiative and automated hierarchical systems for reducing planning effort and increasing plan evaluation measures. However they have still not yet been used in project planning, in part because the relationship between project planning and hierarchial planning is not established. This seminar deals with identifying this relationship and explaining how project planning representations called Work Breakdown Structures(WBS) are similar to plan representations employed by hierarchical planners. This similarity is exploited and then well-known hierarchical planning techniques are applied.
INTRODUCTION
A Project is defined as an endeavor to create a unique product or to deliver a unique service. Examples of projects could be dam-constructions or Enterprise-wide Software system development.
To be successful these projects must be managed. Project management typically includes Planning and Execution sub-processes.
Planning can comprise the following knowledge/work activities and decisions.
1. Creating a work-breakdown structure(WBS): The planner identifies and establishes a hierarchically organized collection of tasks that enables the delivery of required goods and services.
2. Identifying/Incorporating task dependencies: The planner identifies task dependencies and schedules tasks accordingly.
3. Estimating task and Project durations: The planner estimates the time required to complete each task and uses the task-dependencies in WBS to estimate overall project duration.
4. Resource identification, estimation and allocation: The planner identifies the types of resources required by each task, allocates them to each task in the WBS, and estimates the rates of resource consumption
5. Estimate overall project costs or budget: The planner estimates the cost of resources consumed, compiles an overall project cost and often derives a scheduled cash flow.
6. Estimates uncertainties and Risks associated with tasks, their schedules and resources.
Execution can include the following activities:
1. Acquiring and organizing the resources.
2. Performing the task.
3. Monitoring the task status and comparing it with expected execution status to identify deviations.
4. Replanning or adjusting the plan as needed to meet the project objectives.
Several software packages for project management are commercially available. These include MS Project™ (Microsoft), SureTrak™ (Primavera Systems Inc.), and Autoplan™ (Digital Tools Inc.). These packages help a planner to record his/her plan and its associated decisions in a WBS format. WBS is a vital input to the plan execution process. These packages also assist a project manager in plan execution. However, they do not assist a planner in the complex and knowledge intensive task of plan creation and development.
The primary planning activity of a project involves creating a WBS for it, which requires decomposing the project’s tasks into manageable work units. This process requires significant domain knowledge and experience. For example, a software project manager who needs to deliver a real time chemical process control system must employ significant knowledge of real-time software development processes combined with his/her experiences in chemical process control. The complex interdependencies between task and domain knowledge make creating the WBS a difficult task.
By assisting project planners in the creation of work breakdown structures, intelligent planning systems could significantly expedite the planning process and increase its chances of success. Our proposal is based on two areas of research: (1) Automated hierarchical planning systems and (2) knowledge management (KM), which advocates reusing previous problem solving and decision making experiences to improve organizational processes. Building on this foundation, I present an architecture for knowledge-based project planning. This architecture employs an integrated set of methodologies, including hierarchical plan generation and case retrieval, for reusing experience to support a project planner in the creation of a WBS.
HIERARCHICAL PLANNING
Consider a personal transportation domain where one of the tasks could be to travel from Dharwad to Bangalore. The figure shows a hierarchical plan for this travel task as decomposed into three subtasks: call a taxi, take the taxi from Dharwad to Bangalore, and pay the taxi. In this figure, a solid line with an arrow starting from a task and ending at another task denotes a task-subtask relationship. For instance call a taxi is a subtask of the
travel from Dharwad toBangalore task. A dashed line with an arrow starting at a subtask and ending at another denotes an ordering relation between them, implying that the predecessor subtask at which the dashed line originates must be completed before the successor subtask at which the dashed line terminates