Improving Shared Understanding in Multilevel Planning

Planning is a specific example of a problem solving activity that is undertaken across multiple (human) collaborative agents. In order to collaborate, these humans need to form a shared understanding of various aspects of the plan, mutual goals, the contexts of the other agents, and the rationale for others decisions and assumptions. Failure to reach this shared understanding can have serious implications to the success of the resulting plan. Currently plans are developed and shared with peer and subordinate units in a static format such as text, diagrams and spreadsheets which do not normally contain any of the reasoning, logic and interdependencies. As a result, the plans are not easy to update and planning tends to take a lot more time than is normally available. Over the last two years we have been developing a representational scheme, called the Collaborative Planning Model (CPM), for capturing plans from planners at different levels of command. In September 2008, the representational power of CPM to support multiple planners collaborating to create a plan, and detect resource conflicts as they arose was evaluated. This paper summarizes the results of the exercise, and discusses areas for further development to make CPM an effective framework for shared understanding in multi-level planning which is expected to improve timely generation of plans.