At T minus six seconds, the space shuttle's three main engines ignite and pilots prepare for launch. At T minus three seconds, the main engines reach 90 percent of thrust. At T minus zero, flight computers ignite two of the world's largest segmented boosters. Boasting a combined thrust of nearly six million pounds, these leviathan motors heave the shuttle from the earth and hurl it into orbit.

As you might guess, these 149 foot tall, 12 foot diameter, 1.3 million pound (when loaded) boosters are quite expensive to build and maintain. To combat the high costs associated with the boosters, our engineers designed the rocket motor segments to be recovered, refurbished, and reused as many as 20 times each. This complex task requires special facilities and equipment.

To transport the booster segments to Utah for servicing, we use special rail cars. These cars are lined with uniquely developed insulation and carry instruments necessary to monitor temperature and climate for each segment. Since we own a limited number of these cars, we must carefully plan every transport to ensure that the cars are available and accessible. To do this, we adopted the use of simulation modeling.

Using ProModel, we created an exact copy of our entire rocket supply chain system to see how variations to the current system would delay or speed up the process. With this model, we can take new launch schedules from NASA and immediately determine the most cost-effective way to meet that schedule. Now we have an alternative to purchasing $600,000 rail cars.

The Process
The transport and recovery process begins at our main facility in northern Utah where we build and service rocket engines and flight hardware. After loading the boosters with fuel, we mount them on special rail cars and ship them via train from Brigham City to Cape Kennedy where they are off-loaded and flown. For safety reasons several engineers in a pullman car accompany the shipment of segments.

After use, the motor segments are recovered from the ocean and the burned out rubber and insulation is removed. Then the segments are loaded onto the rail cars and sent by train back to Ogden for refurbishing. Once refurbished, the segments return to Brigham City where we again load them with propellant. From here, the process starts over.

The Problem
Master schedulers assemble a schedule of how often we need to ship loaded motors to Cape Kennedy and how often we have to bring burned out flight hardware back to Ogden. If we can maintain that schedule, we can maintain the flight rate NASA needs. Our question was how we could make certain we had enough rail cars to support the flight schedule. If we took these rail cars in and out of the shipping fleet for maintenance actions, how could we make sure we always had enough cars available to transport rocket segments?

If schedulers increase the number of flights per year, it poses a potential problem. Since the cost of purchasing additional rail cars is enormous, we must look at other options. Rather than purchase new cars, we could pay huge premiums to one of several different railroads to put the car on a fast train instead of making the boosters wait at several different sites. Alternatives included finding a way to expedite the maintenance activities performed on rail cars or constructing additional service facilities. The problem was that we could not assess the impact of any of these options on the entire system—the supply chain is complex and contains much variation.

The Solution
We created a simulation model with ProModel that allowed schedulers to model travel time, the number and availability of rail and pullman cars, shift schedules, and maintenance activity time.

With the model finished, we set up parameters that the logistics department could change. This enabled them to test possible solutions without worrying about changing the integrity of the model. When they changed model parameters, the simulation provided them with the actual dollar cost of the proposed change. Using ProModel's free run-time capability, we were able to allow the logistics department to use the model on an ongoing basis while we continue to build other applications.

The parameters in this model will allow the logistics department to test potential solutions, including dedicated facilities to service the rail cars, additional workers or shifts, and other changes that may speed up the process. The bottom line is that we are able to save time and money, NASA can meet their flight schedules, and management can feel confident with the proposed changes.

Future Applications
This model will be used on an ongoing basis to test possible ways to improve or completely revamp the supply chain when NASA requires a major flight schedule change. In our next project we will look at synchronized scheduling scenarios to reduce inventory and overall span time.