design software for implementation of random workload reliability model
conduct numerical studies on different types of stress metrics in order to facilitate implementation and integration into products
We have a comprehensive groundbreaking perspective and mathematical model for device/server reliability that allows the user to input his desired metrics for the workload, along with estimation of parameters such as the service rate and arrival rate and arrive at an exact equation for the failure probability under workload. This model doesn't require standard simplifications, so the application covers a broad area of situations. Full background can be found in our publications Stochastic Reliability of a Server under a Random Workload, On Server Efficiency, and this PhD dissertation.