Project Mercury

IBM assembled more than 75 people to work on a mission with enormous stakes. They were tasked to help design Mercury’s worldwide Tracking and Ground Instrumentation System, which would need to not only keep constant tabs on the spacecraft but also simultaneously transmit real-time, raw data to the Goddard Space Center in Greenbelt, Maryland. The life of the astronaut on board, and the success of the mission, would depend on split-second decisions informed by the processed data transmitted from Goddard to Mission Control, in Cape Canaveral, Florida. There would be no room for error.

The team’s engineers divided the load among three large-scale computers: a pair of 7090 transistorized computers, housed in the Goddard Space Flight Center in Greenbelt, Maryland, and a 709 in Bermuda. The 7090 computers were charged with mapping the flight’s trajectory parameters and the real-time position of the spacecraft — and predicting where the craft may go next. The 709 would calculate orbital flight information, chart the craft’s trajectory dynamics during launch and early orbit, and analyze local telemetry and radar data.

A proprietary routing device capable of receiving 1,000 bits of data per second, known as the IBM 7281, enabled communication within the system. As the project progressed, IBM’s role expanded to include the Launch Monitor Subsystem: an additional line of security against launch emergencies. For this, the company installed another 7090, housed at Mission Control, to process and share radar data with Goddard’s two 7090 machines. 

Of course, the system’s success was as dependent upon human expertise as technology. Throughout the duration of the flights, the IBM team received a deluge of data from their computers. Implementing advanced programs and mathematics, developed especially for their system, IBM engineers would translate it all into critical insights for NASA flight controllers. As Arthur Cohen, the mathematician who led IBM’s Project Mercury team, recalled, “We were the central communication hub of Mercury.”

Before the team could declare their system ready, it had to complete a rigorous set of preflight tests, followed by test flights, which had no crew. IBM’s system performed beautifully — keeping the channel of real-time data open throughout the flight and signaling the alarm when anything went awry. (In a subsequent test flight, with Enos — the only chimpanzee and the third hominid to achieve orbit — a malfunction occurred in the reaction control system. This data, however, was transmitted successfully to flight control during the craft’s second orbit. Enos was brought home early, but safely.)

By 1961, Mercury and its supporting systems were deemed ready to power the first manned mission. In May of that year, astronaut Alan Shepard launched from Earth in the Freedom 7. The flight was a success; Shepard had become the first American to travel to space, and the first pilot to manually control the operation of his craft. A year later, John Glenn followed, becoming the first American to orbit the earth in a four-hour, three-orbit flight.

Meanwhile, IBM’s computers in Maryland, Bermuda and Florida were busy keeping tabs on the crafts and the astronauts. One of the system’s most important responsibilities was calculating the path of the spacecraft during launch to assist flight controllers with the “go, no go” decision. Once the crafts were ready to enter Earth’s atmosphere, IBM’s machines and their operators were also in charge of informing tracking stations where the craft might land or when retro-rockets should be fired to return the spacecraft to the desired recovery area. In this way, IBMers remained integral throughout the entire arc of each mission, from launch to landing.