Adaptive Markov Decision Models in Reliability

SB25.1 Inspection Policies for Highly Reliable Systems Stephen J. Herschkorn, Michael Katehakis --- Rutgers Univ., Fac. of Mgmt. & RUTCOR, New Brunswick, NJ 08903 , (herschko@rutcor.rutgers.edu)

We consider a system with known structure and highly reliable¨ components. We discuss issues related to the problem of developing¨ optimal policies for combined inspection and maintenance of such¨ systems.

SB25.2 Optimal Maintenance Policies Under Incomplete Information Michael Katehakis, Apostolos N. Burnetas --- Rutgers Univ., Fac. of Mgmt. & RUTCOR, New Brunswick, NJ 08903 , (mnk@rci.rutgers.edu)

We consider the maintenance/replacement problem of a single machine¨ with unknown failure characteristics. We develop an adaptive¨ maintenance policy which has optimal properties with respect to the¨ rate of increase of the finite horizon costs under easily verifiable¨ recurrence conditions.

SB25.3 Sequential Inspection Under Capacity Constraints Shaohui Zheng, David D. Yao --- HKUST, Dept. of ISMT, Clear Water Bay, Kowloon, , Hong Kong

We study the inspection process in multi-stage batch manufacturing¨ with capacity limits. The rate of defectives are assumed to be¨ random variables with known distribution. We establish the optimal¨ of a policy characterized by a sequence of thresholds with¨ randomization at no more than 2 threshold values.

SB25.4 Queues with Unreliable Servers & Incomplete Information Apostolos N. Burnetas, Gia-Shie Liu --- Case Western Reserve Univ., Dept. of OR, 10900 Euclid Ave., Cleveland, OH 44106-7235, (atb4@po.cwru.edu)

We consider the group replacement problem for the servers of an¨ M/M/c queueing system under incomplete information on system¨ characteristics. We present the structure of optimal replacement¨ that do or do not depend on the number of customers in the system,¨ and discuss cost comparisons between the 2 cases.