How GEODSS Works

GEODSS telescopes operate in the sidereal-track mode, which means they move at a rate and direction that makes the stars appear motionless in their field of view. To locate deep-space objects, their sensitive cameras take a sequence of rapid electronic "snapshots." If the snapshots were displayed on a screen, stars would appear as dots, and satellites as streaks. Using data from the streaks, the computers calculate the objects' angular positions. The objects are primarily at ranges between 3,000 and 22,000 miles (deep space).

"This is an extremely cost-effective way to catalog these objects," says GEODSS project leader David Potter. "Telescopes let us do long-range detection inexpensively. Telescopes can detect sunlight reflections off of deep-space objects. Conversely, ground-based radars rely on costly energy that must be transmitted from earth and reflected back from objects in deep space with enough power to be detected."

The installation at Diego Garcia, showing the three telescope towers.

The GEODSS system is now being upgraded under two programs. One program upgrades the computer systems (including Y2K testing) and adds a facility that will serve as a command center for the three telescope sites and an interface to the Space Control Center. Another program is refurbishing the telescopes and their mounts. MITRE is providing system engineering support to both programs.

Finding the Problem

"If the company knows it, the customer can know it too," MITRE's President Vic DeMarines said at a recent all-employee meeting. "A MITRE team can reach back to a huge storehouse of specialized knowledge, and this is one of the benefits we offer our customers."

MITRE can look at a problem, recognize it as a control-systems problem, and reach back for a control-systems expert to deal with it. Typically, a contractor who manufactures a portion of a system has a narrow, specialized knowledge -- and the acquisition team needs members with a broader systems perspective. MITRE's system-wide perspective adds value to the team.

In GEODSS, this system-wide perspective proved essential when problems arose with the telescope Mount Precision Angular Control System (MPACS). The equipment was the critical link between a new computer and the existing telescope, and it was causing difficulty for the contractor. MITRE Sensors Systems Lead Engineer Walter Faccenda called it "the brick wall that kept the final pieces of the system from coming together."

Describing this problem, Faccenda says, "A lot of problems have little symptoms that are somewhat universal. With experience, you know how to recognize those problems right away. For systems problems, it's a matter of recognizing that virtually nothing works unto itself, that everything is interrelated, so the source of the problem may not be where you see the symptoms.

"On GEODSS, we heard, 'The MPACS keeps failing; it's unreliable.' As a solution, new equipment was proposed, which would have required additional funding. As we heard of the problems, we evaluated their effect on performance, particularly across interfaces with the rest of the system."

Faccenda tapped into MITRE's pool of expertise and brought in someone who had years of experience in control systems. "To me, it acted like a control systems problem," he said. "I knew a first-rate control systems expert at MITRE, so we brought him (Lead Staff Dr. Mike Hough) into the team. We had a general sense that it was an interface problem, and he got right to specifics: 'This looks like something isn't in sync, and here are the kind of things that suggest that sort of problem.' This led to an interface investigation that revealed a misimplementation by the contractor. Straightening out the problem avoided the cost of a new MPACS."

End-to-End Analysis

Faccenda says, "We can look at a system and perform an end-to-end analysis that evaluates physical phenomena in the sensor's operational environment. Evaluating real-world physics is equally as important as system performance. To determine the signal strength and natural background noise presented to the telescope, we start by considering the solar illumination on the satellite and the satellite's reflectivity. We quantify the effects of the atmosphere on the satellite's reflected light as well as light from sources other than the satellite. We evaluate the telescope and camera's performance as it picks up the incoming light, how the camera transduces the light into an electronic signal, and how instrumentation adds noise. We analyze and evaluate the algorithms used to gather the signal information from the raw data. In turn, we evaluate how efficiently the computers run the algorithms."

Critical support from MITRE staff in Colorado Springs ensured that GEODSS' external interfaces were up to speed. Senior Staff Reid Slaughter says, "MITRE was instrumental in getting the new hardware and software integrated into the Integrated Tactical Warning and Attack Assessment (ITW/AA) system, including implementing interfaces with several existing systems, known in the Space Command community by their acronyms: SPADOC, CSSR, CMP, ASCC, and the Offutt MPDS. We were also the driving force behind the identification and development of new concepts of operations and operational procedures for the Air Force organizations that actually use the new systems."

"Most importantly, as systems engineers with understanding of end-to-end performance, we can identify cost-efficient system improvements. Nobody else is in a position to do all of that."

Continuity over Time

Potter gives an example involving the telescope cameras of how MITRE's perspective provides continuity over time. He says, "We realized the Ebsicon cameras at the back of the telescope were running out of inventory. There's nothing available today that can do what those tubes can do. These are old vacuum tubes, customized in a very specific way for this telescope to work at low-light level as low-noise TV-like cameras. (TV implies a certain number of lines and pixels, and these don't adhere to the definition of TV as you see it, but functionally operates like a TV.)

"The Ebsicon camera's capability is very impressive even now. The problem is that nobody can build a tube like it. Technology has gone to solid state, and the supporting technology and industrial infrastructure needed to produce these vacuum tubes is no longer there.

"We put together a cost analysis for a replacement Charge-Coupled Device (CCD) camera as a modern solution with potential for significant improvement in sensitivity performance. Lincoln Laboratory had built a CCD chip for an earlier program, so we knew a suitable chip could be made. It's a very specialized chip: excellent sensitivity, high frame rate, and very low noise.

"We developed a very detailed cost estimate to build and install new CCDs. Faccenda was the lead on this. He worked with MITRE's Economic and Decision Analysis Center to come up with a systems engineering cost estimate that we provided to the program office here at the Electronic Systems Center. The program to provide CCDs has been approved and should start next year.

"Gaining an understanding of all the nuances of Ebsicons and CCDs, and how to optimize their performance in the system, has been a very satisfying process. Therein lies the systems engineering. As you work with these things longer and longer, you recognize the strengths of certain architectures.

"At a high level, we can look at the development of a technology over time and see if it's on a reasonable schedule and cost, and add our management expertise. This was a good sideline effort to our other work that added high value and impact.

"I have known four ESC program managers since I started on this effort. Faccenda knows five. And there have been several contractor managers during the same period of time. The contractor team changed substantially -- today, there are only a few staff still working since the beginning of the effort. MITRE provided a lot of continuity with their system engineering perspective and expertise in critical areas of software development, data processing, and even optics applications."

Darleen Druyun, Principal Deputy Assistant Secretary of the Air Force for Acquisition and Management has said that MITRE offers a "depth, breadth, and continuity of knowledge over time."

At a critical point in the program, former MITRE Project Leader Eric Koerner led a "red team" review, which analyzes the contractor's risks to completing the program. Describing this situation, Koerner said, "The user community was ready to cancel the program due to cost and schedule overruns. The contractor was really struggling, and their staff turnover was extremely high. We levied MITRE's systems expertise and worked in a team environment with the contractor to identify a realistic plan and course of action.

"MITRE's Software Lead Phil Huggins was a key player in working with the contractor to develop a realistic schedule of achievable milestones. Huggins drew upon MITRE experience and then worked with the contractor to implement the new processes. The user community reluctantly allowed the program to continue, and now it is nearing a successful completion."

GEODSS Part 2 Depth and Breadth of Knowledge

In working with complex problems, Faccenda says, "You have to recognize where to go after things; that's the key, not to troubleshoot things that aren't problems, and to troubleshoot all the things that are problems in a quicker way. You have to find the real source of the problem and not superficially fix some attribute and not realize there is some underlying problem.

"We've learned to look very carefully at assumptions. For example, streaked images were assumed to be a 'focus problem.' The camera goes through a series of steps, automatically optimizing the contrast, and determining an optimal pointing angle to center the satellite in the field of view. As it goes through those steps, it also performs a zoom where magnification changes very rapidly. Usually, imagery isn't seen because the shutter is closed. However, in this instance the zoomed stars produced a bunch of radial streaks -- a spoke-like image. Instantly, I knew that it wasn't a focus problem, but a stuck shutter.

"We have to evaluate the impact of a 'problem.' Is it a big problem? Small problem? No problem at all? We spend more time talking about things like that. We ask the contractor some very specific questions, and depending on what the answer is, we might say, 'We think you have a problem. Here's where we think it is.' Or we might say, 'Don't even worry about it. Put that out of your mind because that's just an artifact of something else.' We help prioritize the work."

A systems perspective is valuable for evaluating proposed changes, too. "For example," Faccenda says, "on a regular basis, ESC gets proposals to use the GEODSS telescopes to collect different types of information, often from people with optics experience. Usually, their experience is at the academic level. In the case of GEODSS, the optics are designed to accomplish one specific thing, and if you try to use them for something else, it's probably a waste of time and energy.

"One proposal was to use the optics as an infrared telescope. Unfortunately, some of the refractive elements are totally opaque to IR. Another proposal was to use the telescope in an adaptive optics environment. Adaptive optics essentially take the twinkle out of stars -- the aberration induced by the passage of the light through the atmosphere, which is not a homogeneous medium. An adaptive optics system, often called 'rubber mirrors,' can correct the image. The mirror surfaces of the GEODSS telescopes were not designed to image details, so adaptive optics are impractical.

MITRE: An Interesting Place to Work

Working on systems of national importance and providing "solutions that make a difference" make MITRE a rewarding place to work.

If you had worked on the GEODSS program, your assignments might have taken you to the top of Mt. Haleakala on Maui, or to the inside of Cheyenne Mountain in Colorado Springs. Although you would be using your expertise to upgrade an elecro-optical system, your coffee-break conversations might include more fanciful knowledge, such as why, in the middle of the night, hundreds of cows block the road to the GEODSS site on top of Mt. Haleakala, or why North Pacific humpback whales travel thousands of miles from Alaska to breed in Hawaii's warm waters.

Senior Staff Bob Jurgens, the most widely traveled MITRE staff on this program, says that a reward for him in working on a program like GEODSS is the people. "I've spent time at all the sites and have met about everyone working the program -- developers and operators. I've traded chili salsa recipes in Maui, and appreciated the persistence of a Diego Garcia operator who has worked GEODSS since it began in the mid-seventies."

Sometimes the opportunity to make a difference arrives unexpectedly. Faccenda recalls a couple of incidents: "The foolish things you remember... On my second week at MITRE, I was looking at plans for how the contractor planned to support a telescope on its side during transportation (with its 500-pound primary mirror on a 1/16 inch shaft) was very likely to have resulted in catastrophic damage! Immediately, the support design was changed to take advantage of the telescope's inherent strengths while in a vertical orientation."

"On the same system, the requirement was that the system had to be located within 45 degrees of the equator. While sitting in a meeting, I looked at the model and recognized that it couldn't work close to the equator because of a mechanical interference. It was one of those moments when you look up and say, 'That isn't going to work!' It was only usable from 11 to 45 degrees from the equator, even though its optimum location for detecting orbiting deep space objects is at the equator. Consequently, the contractor redesign eliminated the interference. For this instance, the insight had nothing to do with setting up goals and going through the systems engineering process to arrive at a product."

Sometimes "depth, breadth, and continuity of knowledge" show up at unexpected times, in unusual places, and on serendipitous areas of expertise, but the big picture of how MITRE adds value remains very much the same from program to program.

Jurgens states "While working with contractors at Socorro during development, test and evaluation of the current computer upgrade, this unique image came up on the high resolution display while troubleshooting a problem with clustered satellites.

"It's rare to get satellites superimposed on something interesting astronomically speaking. Fortunately, the system had a new capability for storing images in a format developed by NASA for astronomical observatories. I said, 'Save that image!'"

MITRE's graphics department laid out this storyboard for the GEODSS ribbon-cutting ceremony on top of Mt. Haleakala, 21 October 1998, attended by Senator Inouye, General Patrick Gamble, Program Manager Mike Findley, and Reverend Kealahou Alika.

Download the GEODSS Poster [PDF, 1.75MB]

Page last updated: August 31, 1999   |   Top of page