Location
Cocoa Beach, FL
Start Date
7-3-1966 8:00 AM
Description
The imposing challenge of space for management is to direct the most complex and massive engineering effort ever attempted to attain obfectives of the highest national priority. And within the body of new management knowledge and experience produced in these tasks lie methodologies fundamental to the administration of any large project. The purpose of this paper is to present an operations research model for the range scheduling function. These techniques were derived from an analysis of the Eastern Test Range, but they are applicable to many cases of the general large project scheduling problem. The method is illustrated with test cases.
A model is presented that would provide computer assistance to the scheduler. The model processes data in the following way. First, requested start times and slack are used to develop a primary "network" whose nodes are individual subtasks'of range tests and whose arcs define "order" relationships among the tasks. A secondary network is then constructed from the first. "States" of the primary graph—sets of tasks containing all task predecessors—form nodes of the secondary graph; each node is connected by an arc to a predecessor state and to successor states. Construction of arcs among the states is governed by resource requirements and the precedence relationships of the first graph. Any route through the second network is a user-feasible, conflict-free schedule. Management objectives, abstracted into measures of effectiveness, may then be used to select the "optimum" schedule. In particular, the most compact schedulethe feasible schedule with highest utilization—is given by the shortest path through the graph and this objective, in conjunction with user feasibility, is proposed as the selection criteria.
The purpose of the range scheduling model is to produce an assignment of starting times to range jobs such that, based on the planned execution times, there exists in the schedule no conflicts over major range facilities. Both weekly scheduling and the real-time rescheduling task require this capability, so that the mechanics of both problems are satisfied with the same model, though variations for special effects may be needed. Although a new technique is proposed, the end product exists today, and management procedures, familiar to users and suited to the role of ETR is the space program, remain unchanged.
Automation In Range Scheduling
Cocoa Beach, FL
The imposing challenge of space for management is to direct the most complex and massive engineering effort ever attempted to attain obfectives of the highest national priority. And within the body of new management knowledge and experience produced in these tasks lie methodologies fundamental to the administration of any large project. The purpose of this paper is to present an operations research model for the range scheduling function. These techniques were derived from an analysis of the Eastern Test Range, but they are applicable to many cases of the general large project scheduling problem. The method is illustrated with test cases.
A model is presented that would provide computer assistance to the scheduler. The model processes data in the following way. First, requested start times and slack are used to develop a primary "network" whose nodes are individual subtasks'of range tests and whose arcs define "order" relationships among the tasks. A secondary network is then constructed from the first. "States" of the primary graph—sets of tasks containing all task predecessors—form nodes of the secondary graph; each node is connected by an arc to a predecessor state and to successor states. Construction of arcs among the states is governed by resource requirements and the precedence relationships of the first graph. Any route through the second network is a user-feasible, conflict-free schedule. Management objectives, abstracted into measures of effectiveness, may then be used to select the "optimum" schedule. In particular, the most compact schedulethe feasible schedule with highest utilization—is given by the shortest path through the graph and this objective, in conjunction with user feasibility, is proposed as the selection criteria.
The purpose of the range scheduling model is to produce an assignment of starting times to range jobs such that, based on the planned execution times, there exists in the schedule no conflicts over major range facilities. Both weekly scheduling and the real-time rescheduling task require this capability, so that the mechanics of both problems are satisfied with the same model, though variations for special effects may be needed. Although a new technique is proposed, the end product exists today, and management procedures, familiar to users and suited to the role of ETR is the space program, remain unchanged.
