incomplete, and they are categorized by mission type. A successful mission means that it was able to fly the mission duration without a critical part failure, otherwise the mission is aborted. An incomplete mission is one that was not assigned due to the lack of aircraft which could fly that particular mission.
3.6. Incorporating Randomness
Incorporating randomness in a simulation is an important feature in quantifying the uncertainty in the process. Most of the values used in the simulation can be defined as a random variable. Some of these variables include mission length, part life time, replacement time of failed parts on the vehicle, repair time of parts, and transit times between the part repair depot and inventory location.
3.7. Model Features
Currently, VE-OPS has several features to perform sensitivity studies and to analyze the impact of different strategies. To study the impacts of the constraints imposed by workers and workstations, the model can limit the number of hours the workers can work as well as the total number of workers and workstations. Depending on the vehicle, it can have different modes of operation or missions it can perform. For example, a multi-role fighter can be used for an air-to-air engagement, an air-to-surface mission or an intelligence, surveillance and reconnaissance mission as well as training operations. The vehicle can also be in partial mission capable status where it can perform some of the mission types but not all. The current model can handle a table of parts with different characteristics such as part life, time to replace a part on a vehicle, time to repair the broken part, and time to transport the part between the depot and local inventory. The amount of parts in the inventory can also be set as an input to simulate a shortage of parts. Some vehicle repairs can be performed in parallel while others must be performed in series, and to capture this constraint, the parts can be designated into repair groups where each part in the group can be replaced in parallel. Some parts are critical to certain missions but not critical to flight, and if these parts fail, the vehicle can still perform other missions and complete them. Because the simulation is object-oriented, multiple instances of the model can be connected to simulate contingency events and to study the transitional behavior. For example, air cargo transport vehicles are important for humanitarian aid in places that are difficult to reach, and it is useful to understand how the vehicle sustainment requirements change when supporting these missions. The transition between the normal and contingency operation can be simulated by creating one model for the normal operation and another for the contingency operation, and the two models are connected so that the vehicles and parts can be transferred from one model to the other. To enable this analysis, there are two additional process classes that move the vehicles and distribute spare parts between inventories.
3.8. Future Development
This simulation model environment VE-OPS supports future research in the area of maintenance and logistics. One area of interest is to investigate repair policies of partial mission capable (PMC) vehicles. A vehicle that can perform multiple types of missions does not necessarily need to be repaired to full mission capable (FMC) status depending on the future mission requirements and the fleet-wide capability. It may be more beneficial to the overall mission effectiveness if it forgoes the unscheduled maintenance for certain mission specific components until there is a more opportunistic time in the future. Another area is to research part allocation schemes between the central and forward supply depots. There are several additions to VE-OPS that will enhance its analysis capabilities and widen its application areas. Some of these include the ability to cannibalize vehicles for parts, an option for three-level maintenance, a prognostics health management (PHM) system, and a supply chain network for spares. Another goal is to create process classes with different levels of fidelity and complexity so that users can choose and customize features of their models. Not all studies need the maximum fidelity in all segments of the simulation. It would be better if the model is composed of modular segments, with each having different variants. By making the segments modular, the