NASA Bids Farewell to the Mission Operations Computer

From the beginning of the Mission Control Center in Houston, we have relied

upon ‘centralized’ ground based computers to process and display telemetry,

format commands, and perform trajectory processing. This has always been done with a

IBM style mainframe computer starting with five IBM 7094’s in the Gemini

program. They were replaced with IBM 360/75’s for Apollo, IBM 3081/3084’s

during the middle part of the Shuttle era and IBM ESA 9000’s in the recent

Shuttle era. These were known as the Mission Operations Computer (MOC).

Over the last several years, various functions of the MOC have been offloaded

to the new ‘distributed’ architecture in the MCC. This started with various

Telemetry Processing and Display functions as the new DEC/Alpha workstations

gave flight controllers the capability to better format telemetry from the

spacecraft and organize it into a much more concise, human interface. During

a transition period that lasted several years, flight controllers could pull

telemetry data from the MOC or directly off the MCC distributed network.

A few years ago, a “Command Server” capability was established, using MCC

workstations to perform the Command authentication and control functions of

the MOC. Although this had some interesting times during it’s testing and

first year of operations, it has turned into a very stable platform. The

command functions on the MOC have not been used for nearly two years.

The only remaining function on the MOC was trajectory processing. This has

always been the most CPU-intensive function and several attempts to port

this function to a workstation/server environment, starting back in the late 1980’s,

were unsuccessful. Over the past few years a new effort has been made

leading to what is now known as the “Trajectory Server”. Extensive testing,

along with lessons learned from previous attempts, have demonstrated these systems

to be very robust.

Shortly after launch on STS-110, we transitioned support

from the MOC to the Trajectory Server, which supported all orbit and entry

phases of the mission without significant anomalies. STS-111 was supported

from ascent through entry on the Trajectory Servers as well – again without

any significant issues. In both cases, the MOC was available as a backup in

the event of a problem.

The plan was to continue supporting the next two missions (STS-107 and

STS-113) with the MOC as a backup, then decommission it by the end of the

fiscal year (September 30). Due to the mission delays that have pushed the

next flight into the next fiscal year, along with the unqualified success of

the Trajectory Server on two flights, as well as many simulations and tests,

the decision was made to accelerate the schedule to decommission the MOC.

This afternoon, in the large computer room on the first floor of the MCC

that has been the home of the MOC from the early 1960’s, a large crowd of

Flight Controllers, Flight Directors, Engineers, Technicians, and management

witnessed the final power down of the MOC. Missing from this event,

unfortunately, were the computer operators and supervisors responsible for

actually running the MOC.

They were laid off two weeks ago.