Following two successful launches, the first on August 20, 1975, and the second on September 9, 1975, we looked forward to a vacation. It seemed that the hard work was over, at least for the next few months. The cameras were on their way to Mars; we were freed from any more hardware decisions. Commands for the first pictures were preprogrammed in the Lander computers. What more could we do than sit back and wait during the 10-month journey to Mars? We soon found out.
The Viking plan called for a mission operations team of more than 800 persons to be in residence at the Jet Propulsion Laboratory in Pasadena, Calif., directing the spacecraft once they arrived at Mars. The majority of these 800 persons would be visitors at JPL. Large groups arrived from Martin Marietta and from Langley. Representatives from many subcontracting companies were on hand to monitor the performance of particular instruments. The scientists, of course, came from many institutions throughout the country.
It was obvious that there were problems, technical and social, in assuring that this large group worked together effectively. If everyone dropped in at JPL just a few days before the landing, chaos would result.
The proposed solution was yet another series of tests. All of us were encouraged to be in residence at JPL, starting in January 1976. I greeted the news with enthusiasm. It seemed a delightful way to escape a New England winter, and surely there would be ample opportunity for travel and relaxation in southern California. In the months that followed my wife pointedly reminded me of my prophetic gifts. Between January and November 1976, we had time for just two weekend trips.
Our first task was strictly bureaucratic. People from many backgrounds had to be blended together in functioning groups. There were the usual office politics-assignment of office space, secretarial services, and the like. Other problems were more intriguing. Scientists especially those working in a university environment-are notoriously independent. They had no objection to being under the nominal control of group leaders from JPL and Martin Marietta, but real control was another matter. With their investigations in the balance, they rebelled against the occasional group leader who attempted to keep all decisions to himself. After a few weeks of job shuffling and personality testing, we finally settled down to the job at hand.
A series of tests was designed to simulate the actual mission operations. To provide a sense of reality, a group of "instructors," working in secrecy, designed a series of mission events that were programmed into the computers. Accordingly, the computer outputs simulated transmissions from a spacecraft on Mars.
The first test, three days in duration, assumed that things were working well. Our only job was to prepare updated command loads to respond to specific situations on Mars. We discovered early a frustrating problem that was to plague us throughout the mission. Any change in the sequence already stored in the Lander computer-even an apparently minor one-required days, if not weeks, to implement. The reasons were several. First, decision making involved a pyramidal structure. All decisions initiated at the bottom of the pyramid required approval at higher tiers, in many cases all the way up to Jim Martin, the Project Manager. Every decision required a meeting. The meetings were seemingly endless, sometimes even overlapping.
A second problem involved safety. To a degree, operation of any one instrument on the spacecraft jeopardized the operation of another instrument. Simultaneous operation might result in a power drain or a computer failure. Less dramatic perils lurked in the background. Designation of a picture somewhat larger than originally planned might utilize all the recorder tape, leaving no room, for example, for some previously planned biology data.
This maze of interconnected events required an incremental approach. A week or so before a command update was scheduled, a candidate sequence of events was determined. Then it was run through a preliminary computer program to detect errors. Inevitably some were found, corrected, and the computer's approval again requested. Eventually passing this hurdle, the sequence of events was translated into the special command words that would be transmitted from giant antennae on Earth to the Lander on Mars. Each final sequence was double checked by using it to operate a test vehicle on Earth, guaranteeing that the commands produced the desired result.
Clearly, nothing like real-time control of the spacecraft on Mars was possible. Even if there were no other constraint, the vast Earth-Mars distance meant that it took about 20 min for a radio signal from Earth, traveling at the speed of light, to reach Mars.
The tests became more challenging when anomalies were introduced. Our instructors would simulate some disaster, leaving it to us to identify and respond to the problem. If our diagnosis was faulty, our corrective surgery clumsy, we were required to repeat the  test. It was like a recurring bad dream come to life. We were back in school, suffering through final exams.
One anomaly test, simulating the period just before landing, was particularly disconcerting. The problem was straightforward, but project engineers seemed unwilling to acknowledge transparent truth. The cameras were scheduled to be turned on, in this way confirming that they were still working after their long trip to Mars. There were two tests to be conducted. First, a small light was turned on inside the camera which flooded all the photodiodes. Although no image could be acquired in this way, the electrical signal recorded by each...
...diode indicated that it was operational. In the second test, a small light bulb installed on the protective post covering the recessed camera window was turned on. In this so-called scan- verification test a conventional, if somewhat drab, image of the circular light was obtained. A single diode was assigned to the scan-verification sequence; other diodes could not be substituted.
In the anomaly test simulating the prelanding checkout, the flooding of all the diodes with the internal light indicated that one of the 12 was not working. That one inoperative diode was the same one used for the scan-verification test. Predictably the scan-verification test produced no data. The solution appeared simple. All we had to do was avoid using the one faulty diode. But, we were asked, how could we guarantee that there was not a double failure? How could we demonstrate that the scan motors had not failed along with the one diode? How could we demonstrate that further operation of the cameras would not endanger the entire Lander? The more we protested, the more complicated became the solutions. Mercifully, the test was ended by a phoned-in bomb threat (not part of the planned simulation). By the time the buildings were evacuated, searched, and normal schedules reestablished, no one could recall just what had happened during the final critical moments of the simulation, when a decision on landing should have been reached. We were depressed. Surely, we told one another, nothing in real life could be as bizarre as these simulated concoctions.
One of the final tests was a pleasant change from our previous problems, regardless of the fact that they were simulated. The landing sequence was reviewed and we sat in front of the television consoles as the first pictures from "Mars" were displayed. Of course, these pictures had been taken at a nearby locality and were appropriately coded to simulate a transmitted picture from the Lander. As we watched the first image appear, an undistinguished sandy surface was revealed. But then, in the midst of all this sand, a symmetrically jointed lobate object, perhaps 15 cm long, appeared (fig. 20). Alan Binder, standing next to me, exclaimed, "That's a trilobite!" (Trilobites are primitive arthropods that thrived in the early Paleozoic seas on Earth, hundreds of millions of years ago. Their external skeleton, with head, body, and tail, creates a distinctive fossil that is much prized by professional paleontologists and amateur rockhounds.)
Without examining the Viking pictures in any detail, we jumped to the conclusion that some lighthearted geologist had salted our martian scene with a terrestrial fossil. At the "press conference" held a few hours later-yes, even the press conferences were simulated-we reported, tongue-in-cheek, that life had been discovered on Mars. Not everyone was amused. Some of the biologists were irritated by our hasty pronouncements. Even though we were playacting, biology was a controversial subject. Other scientists looked at the picture and doubted our identification. The more I studied the elliptical object, the less confident I became. It was clear that our initial identification was intuitive. But what if our intuition was wrong? It would not speak well for our credibility during the actual mission.
With the test completed, we were allowed to see the scene that had been photographed. Our fears were unfounded. There, nestled among the sand ripples, was a magnificent replica of a Cambrian trilobite.