Voyager: Exploration, Space, and the Third Great Age of Discovery

by Stephen J. Pyne

  Over the course of construction the validation procedures racked up 3,500 reports of problems or failures—this out of a machine that had 65,000 individual parts, and complicated interactions among all of them, guided by first-time software. Some of those “parts,” moreover, were themselves electronic composites. (Each Voyager was estimated to have the “electronic circuit complexity” of 2,000 color TV sets.) As launch approached, the pace of testing quickened—provisioning at port has always seemed a frenzy. Program manager John Casani recalled a “lot of crisis activity” during those last few months—“changing radio systems and switching things around,” many of these difficulties “unexpected” and “quite challenging.” In that final rush a major loss of detectors occurred, because the building in which the spacecraft was being tested got painted, and spray (“some hydrocarbons”) was sucked inside the assembly room through the air handler and contaminated the sensors.61

  Not all the testing failures resulted from instruments. Engineers had not fully appreciated the actual stresses from launch, and had not tested Voyager for them, and at one point during launch thought they “might have lost the spacecraft.” What had actually happened was that Voyager’s own self-protection systems had kicked in, confusing controllers, who thought they were witnessing outright failure instead of fail-safe software working as it should. The breakdown lay in not devising the right tests. Before Voyager 1 launched, some minor mechanical adaptations were made to dampen the effects the second time around. The reliance on software, however, suggested an alternative to simple redundancy: Voyager’s onboard computers could be reprogrammed. That meant that not every task had to be replicated in metal, or potential failure anticipated; adjustments could be improvised as needed. Voyager could substitute flexibility and redundancy for size.62

  Raymond Heacock observed that the level of redundancy was such that each Voyager contained “almost two spacecrafts within a single structure.” In truth, the issue went well beyond that, since the Voyagers were triplets, and each did work. The proof test model proved critical when, during final assembly, problems cropped up in VGR-2, and the nominally spare spacecraft had to launch in its place. The on-ground model subsequently helped engineers understand glitches and breakdowns when Voyager 1 had its scan platform stick and then had a capacitor short out, and again when Voyager 2 lost its primary receiver. By experimenting with the grounded spacecraft, it was possible to replicate in the lab what telemetry was reporting from space.63

  Despite breakdowns in some critical components and problems with “scientific instruments,” Heacock noted that there was “essentially no mission loss due to failures.” The prime objectives were met—and more. Redundancy once again proved an almost indispensable index of successful exploration.64

  For that reason as well there were two spacecraft. Launch windows were unforgiving, interplanetary space hostile, and distances impossibly remote. If something went wrong, there was no opportunity to return, or to sail to a nearby port or to temporarily beach on a remote moon for careening or repairs. The simplest solution was to send multiple vessels.

  Almost all the Great Voyages were fleets. Columbus sailed with three vessels, then seventeen (a hapless experiment in colonizing), six, and four; da Gama, Cabral, and Magellan each launched with five. John Davis sailed first with two, then four, and then three. James Cook had only the Endeavour on his first voyage (and nearly foundered on the Great Barrier Reef), then sailed with two vessels on his subsequent circumnavigations. There were of course solitary ships, mostly to the better known North Atlantic. On his first voyage, John Cabot discovered “New Found Land” in the Matthew, and returned to tell the tale. That reconnaissance inspired plans to establish a trading colony and a five-ship voyage, of which one ship crept back to Bristol while the other four were lost at sea.

  Even major expeditions had to accept whatever ships they could claim, very few of which were new or prime. (Da Gama’s voyage mattered sufficiently to warrant the construction of two new ships, but that was an anomaly.) While Magellan was assembling his fleet, the Portuguese consul in Seville reported gleefully that his ships were “very old and patched up,” and that he personally “would not care to sail to the Canaries in such old crates; their ribs are soft as butter.” In the end only one patched-up ship and eighteen men survived (although one, the San Antonio, had deserted earlier). That was far from exceptional. Even repairs required extra materiél, and unlike with Voyager, no spare vessel would be left at port to edify engineers. The Great Voyages went in groups.65

  On the continents, save in extreme environments such as Antarctica, the Sahara, and Greenland, it was possible to live off the land or trade with indigenes, so the demands for redundancy could be reduced. That gain was frequently offset by the need in hostile or truly unknown lands for a military escort; the same was true for fleets. A sensible fraction of exploring armadas was devoted to firepower. The more a mission sought to go beyond simple reconnaissance, the more complex its arrangements, the greater the tendency to replace lone travelers with a corps. The Long Expedition of 1819-21 (the first of five by Stephen Long) took nineteen men up the Platte to the Rocky Mountains, beginning with the first steamboat, the Western Explorer, on the Missouri River. It was, one observer exalted, a veritable cavalcade of the era. “Botanists, mineralogists, chemists, artisans, cultivators, scholars, soldiers; the love of peace, the capacity for war: philosophical apparatus and military supplies; telescopes and cannon, garden seeds and gunpowder; the arts of civil life and the force to defend them—all are seen aboard.”66

  That was the norm. The magnitude of the investment, the uncertainties of the hazards, and the expectations of losses, all trended toward small numbers of costly expeditions rather than swarms of inexpensive ones. There were plenty of exceptions, especially by first-voyaging nations. John Cabot sailed to the New World in the Matthew, a navicula about the size of Columbus’s Niña. Some single ships even ventured around the world, as Francis Drake did in the Golden Hind and William Dampier in the Cygnet, but they were privateers, and many of the famous scientific expeditions during the Second Age involved single vessels—think Thomas Huxley in the Rattlesnake and Charles Darwin in the Beagle. But the latter had well-established ports of call at which to provision, and they were surveying points of navigational interest such as the Strait of Magellan or the Torres Strait rather than blue-water voyaging into the unknown.

  The geography of launch windows also argued for multiple or paired launches. Since the ideal Hohmann trajectories to Mars came every twenty-six months, it became the norm wherever possible to launch two spacecraft at each cycle. Some failed on launch; some failed at Mars; but few cycles passed altogether without some contact. Moreover, even paired spacecraft on the same mission might have incommensurable objectives: it was not possible, for example, to fully survey both Titan and Saturn’s rings in one traverse. In this respect, as in many others, planetary exploration more resembles the voyages of the First Age than it does the Second. Voyager, however, was the last of the paired missions. Double-launching spacecraft had become, in John Casani’s words, “an insurance policy that now we can’t afford to take.” In this, too, it resembles its Iberian predecessors who over time risked more and more with less and less.67

  So there were two Voyagers, as there were two captains in the Corps of Discovery, and two ships on Cook’s second and third voyages. But unlike those others, they would not communicate between themselves. They had separate tasks, and once separated, they would not rejoin.

  VOYAGER AS EXPLORER

  Voyager was an autonomous spacecraft; by most standards, the first. “Autonomous,” however, is a relative concept, and its operational meaning is specific to particular tasks.

  There were things the Voyager spacecraft could do—had to do—that previous spacecraft could not. It had to self-regulate. It had to monitor its machinery, take remedial steps in case of sudden ruptures or emergencies, restart, and from time to time accept new orders. Those needs increa
sed over time. So great was the distance that even at the speed of light, messages to Earth and back would take too long, and so extended was the journey that constant text-messaging with, and micro-managing by, JPL would exhaust the spacecraft’s reserves of energy. Voyager had to tend to itself—identify problems, shut down troubled parts, restart anew. It had to assist ground controllers with adjusting trajectories and transmitting scientific data. It had to execute the split-second maneuvers of close encounters through preprogrammed codes, each rewritten based on past experiences. It had to recognize when its onboard machinery for overseeing such tasks had itself failed or become corrupt. It had, within limits, to exercise self-control.

  The technology that made this possible was the digital computer, which coevolved with the space program. Gordon Moore announced his eponymous law in 1965, the same year that Gary Flandro outlined the prospects for a Grand Tour. The second would depend on the first. Moore’s Law states that the number of transistors that can be placed inexpensively on an integrated circuit increases exponentially—specifically, it doubles roughly every eighteen to twenty-four months. From this derives nearly every metric of computing power. By the time Voyager launched, twelve years after those dual announcements, six doubling periods had passed, or a sixty-four-fold increase in computing capabilities. Voyager was on the cutting edge of working computers, and particularly miniature ones. By the time it encountered Saturn, another doubling period had come and gone, and the personal computer was arriving. Voyager seemed vaguely archaic. When it encountered Neptune ten years later, another five doubling periods had so pumped up computing power that Voyager stood to an average cell phone as a caravel to a Centaur rocket.

  Onboard computers evolved out of sequencers, or preprogrammed counters that triggered a series of operations. Once started, a sequencer performed a series of activities that allowed a spacecraft to disengage from rockets, unfurl, and fly past a planet while recording data and images and transmitting them back. Earth controllers initiated the sequence and uploaded new ones for each phase of the mission. A central computer could oversee several such sequencers. For longer flights, such redundancy was mandatory, and as computing power swelled, programmable computers could replace simple sequencers. Voyager had three such computer systems, each with a backup. During its cruise phases, a “sequence load” was uplinked to the spacecraft every four weeks. The uplinked sequences quickened during encounters. 68

  The Attitude and Articulation Control Subsystem computers assisted with navigation. No ground controller could hope to maintain the constant monitoring that told Voyager where it was in the solar system and how to stabilize its three axes. The A ACS computer could, and it kept the spacecraft’s antenna directed to Earth. The Flight Data Subsystem computers performed analogous tasks for the scientific instruments and the transmission of data. Overseeing both subsystems, along with the programmable sequencers that monitored the spacecraft and piloted its encounters, was the Computer Command Subsystem. The CCS was virtually identical to the computer used for the Viking mission. The FDS had to be constructed uniquely. To conserve power, the FDS and AACS computers could go into sleep mode.69

  Command was distributed, redundant, and capable of hibernation. While it ultimately rested with NASA (through JPL), it routinely resided within Voyager. If Voyager went awry, its Earth-based human controllers would have to reclaim and recharter the spacecraft. If within Voyager one computer failed, its backup would step forward, and if one system failed, another might be able to compensate. The ability to upload new software meant that fresh orders could be rewritten as the mission evolved, particularly if Voyager 1’s success meant a Grand Tour was possible. The capacity to power down when not in use meant the RTG power packs could sustain the long journey.

  Yet autonomy, while necessary, could introduce instabilities of its own. Voyager’s understanding was literal and did not always coincide with that of its human controllers. There were occasions when the nominal controllers were confused and onboard computers had to intervene to correct them. (This happened, for example, during Voyager 1’s launch.) There were also events in which the people forgot to perform tasks or did them incorrectly, and the spacecraft went into backup or safe modes. And there were incidents in which “the computers on board displayed certain traits that seemed almost humanly perverse—and perhaps a little psychotic,” as one staffer observed. The programming was too tight, the instructions too sensitive, the tolerance for the quotidian of flight, even cruising, too fine to accommodate the realities of interplanetary travel. The computers overreacted, a kind of automaton hysteria. Some software therapy in the form of new programming sedated the machine and allowed it to regain equilibrium.70

  Even mutiny was possible. There could be internal arguments among the onboard computers, perhaps temporarily paralyzing decisions. But most spectacularly it could happen as it did to Voyager while en route to Saturn, in which new commands were uploaded, only to be refused on the grounds that they contradicted more fundamental instructions and would result in catastrophe. Such insubordination could not be flogged away or strung up on an instrument boom yardarm. The source of the confusion had to be found, and corrected. In this case Voyager—which made the proper decision—was able to defend itself without a court-martial conducted through the Deep Space Network. 71

  How much autonomy was technically possible? Some, with increasing degrees of self-control possible as physics improved computing. How much was desirable? That question edged into metaphysics. No explorer has been fully autonomous. All have had restraints imposed by their mission or their sponsor, and all have had orders that would allow for surprise, adaptation, and redirection.

  What the robots did not have was a human consciousness and conscience. They knew nothing of the traditional horrors, like ship rats, that had ever accompanied long-voyaging expeditions. They would experience nothing akin to the miseries Magellan’s crew faced as they sailed for “three months and twenty days without taking on board provisions or any other refreshments,” during which they ate “only old biscuit turned to powder, all full of worms and stinking of the urine which the rats had made on it, having eaten the good.” The crew “drank water impure and yellow. We ate also ox hides which were very hard because of the sun, rain, and wind. And we left them four or five days in the sea, then laid them for a short time on embers, and so we ate them. And of the rats, which were sold for half an ecu apiece, some of us could not get enough.” The worst affliction was scurvy, which killed thirty-one of the crew. Other “maladies” crippled another twenty-five or thirty.72

  Yet spacecraft had their own peculiar ailments. Parts broke, mechanical gears became arthritic, instruments lost sight, computers suffered hearing loss. There were leaks. Filters clogged. The critical onboard provisions for Voyager were the hydrazine required by its thrusters and the electrical power derived from the RTGs. Without hydrazine the spacecraft could not maneuver, and without power, it could execute none of its mission or even communicate. It would succumb to robotic scurvy. It would become a ghost ship, at the mercy of solar winds and gravitational currents. Earth, Mars, Venus, and the Moon were littered with wrecked vehicles. The Mars Observer had vanished in space.

  Voyager knew none of this—could not fret over what might cripple it, could not leap into the breach or suddenly freeze in fear, could not imagine failure or triumph. Its designers could, and they sought to program the spacecraft to handle both the expected and the unexpected. But the only way to cope with the myriad possible hazards was to grant Voyager some capacity to respond on its own. To find its own way it needed something like its own will.

  VOYAGER AS GRAND GESTURE

  Still, if Voyager was less than human, it seemed more than a machine. It had symbolic value. In particular, it will likely stand as the grand gesture of the Third Age.

  Such expressions are not the first bold announcements of discovery, which come early, with the shock of first discovery or first expression. Such was the case, for example, wi
th Columbus and Dias, with the campaign to measure the transits of Venus, and with IGY, all of which proclaimed a new age. While that annunciatory first revelation can galvanize the imagination and inspire successors, it comes too soon to capture the still-inchoate features of the times, or it is too tightly bound to a particular people or project. So although the Carreira da India first launched under Vasco da Gama and became the basis for Portugal’s national epic, The Lusíads, Spain could equally point to Columbus’s four voyages; and England to John Cabot’s daring sail to the New Found Lands at the same time; and France, always willing to consider history pliable in the interest of gloire, to Jacques Cartier. The essence of the First Age was, as J. H. Parry put it, the discovery of the sea, not simply the Ocean Sea that lay beyond the classical world or maybe encircled the lesser seas as the Styx encircled Hades, but the astonishing way in which they all connect and flow one into another. The enduring symbol of the age would highlight that fact.73

  Grand gestures seem to follow twenty to thirty years later, after exploration and its sustaining culture have worked through the full terms of engagement, as first discovery becomes full discovery. They are the moments of exploring that more than any other capture the general imagination, that fuse place, time, discovery, and yearning in ways that seem to speak to an era’s sense of itself. If they do not inform an age, they do display the vital attributes of an age as nothing else can.