We model the hierarchical and decentralized nature of product transitions using a mixed-integer bilevel program with two followers, a manufacturing unit and an engineering unit. The leader, corporate management, seeks to maximize revenue over a finite planning horizon. The manufacturing unit uses factory capacity to satisfy the demand for current products. The demand for new products, however, cannot be fulfilled until the engineering unit completes their development, which, in turn, requires factory capacity for prototype fabrication. We model this interdependency between the engineering and manufacturing units as a generalized Nash equilibrium game at the lower level of the proposed bilevel model. We present a reformulation where the interdependency between the followers is resolved through the leader’s coordination, and we derive a solution method based on constraint and column generation. Our computational experiments show that the proposed approach can solve realistic instances to optimality in a reasonable time. We provide managerial insights into how the allocation of decision authority between corporate leadership and functional units affects the objective function performance. This paper presents the first exact solution algorithm to mixed-integer bilevel programs with interdependent followers, providing a flexible framework to study decentralized, hierarchical decision-making problems.

The scheduling of jobs and resources In a manufacturing environment is important because o f its basic impact on production costs, but is difficult because of the problems of combinatorial complexity and executwnal uncertainty. Scheduling suffers from combinatorial complexity because there are a very large number of schedules which can be generated f o r a set of jobs and resources, but there is no good way to choose between the options prior to execution. Scheduling is complicated by executional uncertainty in that unforeseeable events will almost certainly occur to disrupt any particular schedule once execution commences. This paper describes a novel approach to scheduling which overcomes these two difficulties. Implementation of the approach requires a new representation for schedules and techniques for knowledge-based reasoning about this representation. The issues involved in performing the knowledge-based reasoning are described and illustrated by examples drawn from the domain of robotic assembly.

The collision detection problem is encountered whenever objects are to be manipulated in the real world. Efficient solutions to this problem must intelligently decide when in time and where in space to check for interference among the objects. Complete solutions must guarantee that no collision can be overlooked. Useful solutions should also report the precise space/time points at which colliding objects begin and end contact. An algorithm based on velocity and distance bounds is described which is both useful and complete. Experiments are presented which measure the efficiency of the algorithm on a variety of test cases. The algorithm is contrasted with existing collision detection methods.

Impressive strides have been made in dealing with the spatial complexity of robotic assembly tasks. Unfortunately, advances in dealing with temporal complexity have not kept pace. It is proposed that one reason for this deficiency is the unnecessary confounding of planning and scheduling. These two activities are differentiated on the basis of knowledge required/knowledge available at robot programming time. It is suggested that giving the robot the ability to reason opportunistically over knowledge of part availability at run time is a practical, efficient way to streamline assembly tasks. Initial experimental results are presented to substantiate this conclusion.