It so happens the Atlas 5-meter fairing is just big enough to fully encapsulate an Orion crew module. Thus was born the first MLAS concept: we would procure an Atlas fairing and Orion CM mockup and launch them on some combination of aft-mounted solid motors. Several missile programs had readily available designs we could cannibalize; some even came equipped with Thrust Vector Control (TVC), an added bonus. When we started down this path we envisioned four clusters of two motors arranged around Orion's Service Module (SM). One motor of each pair would have a high-thrust, short-burn profile suitable for accelerating the CM and fairing off the pad. The second motor in each pair would fire at burnout and provide a lower-thrust, long-duration impulse to complete the flight trajectory. If we could find a suitable motor with TVC, it could even be used to fly the turnaround maneuver and orient the CM for separation; if not, we would could use parachutes to accomplish the same thing.
At about this point, I started thinking about models .... I had this piece of 6" phenolic tubing left over from another project and a drawing of an Atlas 5 fairing. I had a lathe, too, but didn't want to eat a lot of balsa and time learning to use it again, so for the nose cone I turned to Gordon Agnello, aka "Sandman" of Roachwerks Little Joe II fame. Gordon rapidly produced a beautiful cone, all pre-primed and sealed for me. I added 8 24 mm motors, clustered as we proposed clustering the MLAS motors (but canted through the vehicle centerline) and 4 Nike-style built-up fins. Thus was born the original µMLAS, shown on the left in this photo. It made a great show-and-tell model, quite useful for demonstrating the concept to folks as we set about building a real MLAS. And plopping it down on the conference room table that first day was certainly entertaining.