North America

Standby 1 - 2017 ARS North America

These presentation materials will be distributed with the Symposium proceedings, and the author will be available to present if there is a cancellation.

Auto Injector Dispense Time Modeling and Reliability Estimation Using Monte Carlo Simulation

Auto injectors are widely used for drug delivery. Long dispense time could be one of causes for delayed delivery of drug, which is among complaint categories with a high complaint rate. To reduce complaints due to long dispense time, it is of great value to have answers to the following questions: (1) what design parameters have big impact on dispense time; (2) how dispense time is sensitive to these design parameters; (3) what is the probability for the dispense time to be greater than pre-defined threshold values, which is the unreliability defined in this presentation; and (4) how the unreliability is sensitive to design parameters. In this presentation we show reliability methods and techniques to get answers to these questions. It turns out that the unreliability is more sensitive to design parameters than the nominal dispense time, which confirms the importance to look at distribution and reliability instead of only nominal values.

Key Words: Design for Reliability, Monte Carlo Simulation, Sensitivity Analysis, Normal Distribution, Lognormal Distribution, Truncated Distribution, Sampling

Zhibin Tan and Sherwin Shang