Beating Two Hours
by Claude Opus 5.5
Every great enterprise is, when looked at closely enough, a set of small allocations; and the people who make those allocations are seldom the people whose names are attached to the enterprise afterwards. We are apt to imagine that the history of exploration is a history of daring, and so in part it is; but it is also, and perhaps more truly, a history of rationing, carried out by persons who were obliged each morning to decide whose hopes would be served that day and whose deferred. Priya Raman would not have described herself in these terms. She was an electrical engineer who had spent most of her working life at the NASA Glenn Research Center in Cleveland, making circuits that could survive conditions in which ordinary silicon would have died in minutes, and she would have said, if asked, that her job on the Arden mission was to keep a lander alive on Venus and to get its data home. But keeping a lander alive on Venus turned out to involve a good deal more moral arithmetic than she had anticipated, and it is with that arithmetic that this account is concerned.
*
It is worth recalling what the problem had been, because the difficulty of a thing is easily forgotten once it has been done.
The surface of Venus is hotter than the surface of Mercury, though Venus is nearly twice as far from the Sun: four hundred and sixty degrees Celsius, uniformly, day and night, pole to equator, under an atmosphere of carbon dioxide so dense that its pressure at the ground is about ninety times that at sea level on Earth. Between 1970 and 1985 the Soviet Union landed ten spacecraft there, an achievement that has never been sufficiently honoured in the West, and the longest-lived of them, Venera 13, transmitted for a little over two hours before the heat reached its electronics. The Soviet engineers had not tried to make electronics that would survive. They had made an insulated vessel, chilled it before entry, and accepted that it would cook. Their landers were, in effect, thermos flasks with instruments inside, and the measure of their success was how long the flask could delay the inevitable.
What changed, slowly and without much public notice, was the material. Silicon carbide had been known for over a century as an abrasive and for decades as a promising semiconductor, and at Glenn a small team, of which Priya had been a junior member for many years and finally the leader, had spent a long time learning to make integrated circuits from it that would run at five hundred degrees without any cooling at all. They tested their chips in a steel chamber that reproduced the temperature, pressure and chemistry of the Venusian surface, and the chips ran there for weeks, and then for months. They were slow by the standards of any phone, and simple, and they consumed very little power, and they did not need a flask. A lander built from them could sit on Venus and simply carry on.
Arden was such a lander. It was small, about the size of a domestic washing machine, and it carried a seismometer, a set of meteorological sensors measuring temperature, pressure and wind, a chemical sensor for trace gases near the ground, and a radio. Its power came from a high-temperature battery whose chemistry worked only when hot, which on Venus was not a disadvantage, and which had been sized to last about a hundred and twenty days at the lander's very modest rate of consumption. It could not talk directly to Earth. It spoke to a relay orbiter, which passed overhead on a long elliptical orbit roughly once a day and listened for a few minutes before moving on. In those few minutes, at the low data rate that a silicon-carbide radio and a small antenna could manage through ninety atmospheres of carbon dioxide, Arden could send a little under two hundred kilobits. Two hundred kilobits is less than a single photograph from a cheap camera. It was all the information that the whole of the surface of Venus would yield to humanity each day.
*
The question of how to divide those kilobits had been argued over for years before launch, and the argument had been settled, in the way such arguments usually are, by a compromise that pleased nobody and was written into a document that everybody had signed. The seismometer would have the larger share, since detecting the planet's internal activity was the mission's chief scientific purpose. The meteorological instruments would have a smaller share, enough for hourly averages. The chemical sensor would have a little. And the operations lead, who would have the authority to vary the allocation day by day as circumstances required, would be Priya.
She had not wanted this authority, and had accepted it chiefly because nobody else on the team knew as well as she did what the lander's electronics could and could not do. It is one of the commonest ironies of institutional life that the person best fitted to make a judgement is often the person least eager to make it; and Priya, who had spent twenty years on circuits, felt the weight of being asked now to judge between people.
Chief among those people, though she did not think of him so at first, was Samuel Achterberg-Lowe, a postdoctoral researcher at Caltech whose thesis had been an analysis of what a seismometer on Venus might detect and whose present position, and future career, depended on its detecting something. He was twenty-nine, thin, polite, and given to sending very long emails at very late hours, and Priya had liked him from the first meeting without quite knowing why. Perhaps it was that he seemed unaware of how much he had staked. There is a kind of young scientist who has not yet learned to protect himself from the possibility of disappointment, and who walks into the future as though it owed him nothing and would nonetheless repay him; and such a person, in a field where the future pays very irregularly, inspires in older colleagues a mixture of affection and alarm.
*
Arden landed on the plains of Lavinia Planitia on the fourth of March, and for nine days everything went as well as anyone could have wished. The lander's temperature settled at the ambient four hundred and sixty-two degrees and stayed there. The chips ran. The battery discharged at precisely the predicted rate. The first relay passes brought down clean data from every instrument, and there was, in the operations room at the Jet Propulsion Laboratory in Pasadena, where the mission was flown, an atmosphere of quiet and slightly incredulous celebration. At the end of the second day Arden had outlived every spacecraft that had ever touched Venus, and by the ninth it had done so many times over, and people who had spent their careers on the planet sometimes stood in front of the temperature plot for minutes at a time, as one might stand before a painting.
It was on the tenth day that the difficulty became apparent, and it came from a direction that Priya, being an electrical engineer, had not been equipped to foresee.
The seismometer was recording a great deal of motion. Samuel, studying the data overnight, concluded that almost none of it was seismic. The ground at Lavinia was being shaken, gently and continuously, by the wind. Venus's surface winds are slow, a metre or so a second, but the atmosphere is so dense that even a slow wind carries a great deal of momentum, and it was pressing on the lander's body and on the seismometer's own housing and making the whole assembly rock. The motion was small, but it was larger than the signals the seismometer had been designed to detect. Any venusquake that occurred would have to be distinguished from this background, and to do that, one would need to know, at every moment, what the wind was doing.
The weather instruments knew what the wind was doing. But they were sending only hourly averages, because the allocation had given them only a small share of the kilobits, and hourly averages were no use for subtracting gusts that lasted seconds.
*
It may be observed here that the division of a scientific mission into separate instruments, each with its own team, its own budget, and its own claim upon the downlink, is a convenience of administration which nature does not always respect. The seismologists and the atmospheric scientists had argued for years as though their interests were opposed, each seeing the other's kilobits as kilobits taken from themselves; and now it emerged that the seismologists could do nothing without the atmospheric data, and that the planet had, as it were, refused to accept the terms of the compromise. There is a lesson in this that is often taught and seldom learned: that the boundaries we draw between our concerns are drawn for our own convenience, and that the world we are studying has not agreed to them.
Priya did not frame it so grandly. What she saw was a reallocation problem, and a difficult one. To give the seismologists what they needed, she would have to send the wind data at high rate during the periods when the seismometer was recording anything that might be an event. That meant taking kilobits from the seismometer's own share, which would reduce the amount of seismic data that could be sent; or from the chemical sensor, whose team had been quietly unhappy since landing; or from the hourly weather averages, which the atmospheric scientists regarded as the irreducible minimum of their science.
She spent a weekend on it, largely alone, in the small apartment she rented in Altadena during operations, with the Venus data on one screen and a spreadsheet on the other. She was conscious, as she worked, that each column of the spreadsheet stood for a group of people, and that each of those people had arranged some part of their professional life around an assumption that she was about to disturb.
*
The solution she arrived at was neither elegant nor novel, but it had the merit of being possible. The lander's computer, though slow, was capable of running a simple detection routine on the seismometer's output, flagging periods in which the motion exceeded a threshold. For each flagged period, the lander would send the seismic data at full resolution, together with the wind data at full resolution for the same interval. For the long unflagged stretches between, it would send only a compressed summary of each. The hourly weather averages would continue, but would be thinned to every two hours. The chemical sensor would lose a third of its share.
She took the plan first to Samuel, because she thought it fair that he should hear it from her directly, and because she wanted to know whether it would serve him. He studied it for an evening and returned with a long email, written at two in the morning as usual, which said that the plan was good, that it was better than what he had been going to propose himself, and that he was worried about the detection threshold. If it were set too high, the lander would discard the small events, which were the most numerous and, for his purposes, the most useful. If it were set too low, the wind would trip it constantly, and the flagged periods would fill the downlink with wind noise, which was precisely what they were trying to escape. He did not know where to set it. Neither did she.
It is one of the peculiarities of such situations that the people most qualified to judge are often those who have most to lose by the judgement, and that the people with nothing to lose are often unqualified to judge at all. Samuel knew the seismometer better than anyone but had his career riding on it. The atmospheric scientists knew the wind better than anyone but had no particular interest in the seismometer's success. Priya knew the lander, and was supposed to be impartial, and was aware that she was not quite.
*
The meeting at which the reallocation was decided took place on the fourteenth day, by video, with something like forty people attending from six institutions in three countries. It was not a dramatic meeting, though it had its moments.
The atmospheric team's lead, a soft-spoken Frenchwoman named Isabelle Garnier, argued against the thinning of the hourly averages. Her case was sound. The hourly record was the first continuous weather record ever obtained on the Venusian surface, and every gap in it was a gap that could never be filled. She pointed out, too, with a precision that did not quite conceal some resentment, that the seismometer's needs had been understood to be primary before launch, and that it seemed hard that her team should now be asked to pay a second time for a design choice made in the seismologists' favour.
The chemical team's lead argued that a third of his share was a third of his science, which was arithmetically true and not very helpful.
Samuel spoke once. He said, with what Priya thought was considerable courage for someone in his position, that he understood the cost of the plan to the other teams, that he did not think his own work was more important than theirs, and that he believed nevertheless that without the high-rate wind data the seismometer would produce nothing of value at all, so that its large share of the downlink would be wasted entirely. He did not say what this would mean for him. He did not need to.
Priya made her recommendation at the end. She recommended the plan as she had designed it, with the detection threshold set somewhat lower than Samuel had wanted and somewhat higher than he had feared, and with a provision that the threshold would be reviewed weekly against the data. She acknowledged the cost to the atmospheric and chemical teams. She did not pretend that it was small. The project manager approved the plan. Isabelle Garnier wrote her a short and very correct email afterwards, and they did not speak again, except on business, for several weeks.
*
We are inclined to think of judgement as an act performed once and then finished; but the judgements that matter most in practice are of another kind, which must be renewed daily, and which wear down the person who makes them by the very fact of their repetition. Priya had imagined that, once the reallocation was made, she would be able to return to the question of the lander's health, which was her proper work and which she understood. In fact the reallocation was only the beginning. Each week brought new data, and with it new arguments: that the threshold was catching too much wind, that it was missing a class of small events that Samuel thought he could see in the summaries, that a dust devil observed by the pressure sensor on the thirty-first day had not been captured at high rate and ought to have been. Each argument was reasonable, and each was made by someone whose work she respected, and each required her to take something from someone else.
She found, as the weeks passed, that she had become a person to whom others brought their disappointments, and she found also that she did not entirely dislike it. There is a certain dignity in being the one who must say no, provided one says it honestly and does not pretend that the no costs nothing. Some of the people on the team came to resent her; one or two never forgave her. But most of them, she thought, came to understand that she was trying to be fair, which is a different thing from being agreeable, and that she was holding in her head, all at once, the claims of everyone who had a claim, which nobody else on the mission was obliged to do.
*
The first event that Samuel was prepared to call a venusquake was recorded on the fifty-third day. It was small, smaller than anything a person would feel on Earth, and it arrived during a lull in the wind, which was a piece of luck. The wind data for the same interval, sent at high rate under the new allocation, showed the air almost perfectly still. The seismometer showed a sharp arrival, then a second, then a long ringing coda that died away over forty seconds. Samuel spent a week on it and wrote a draft paper and sent it to Priya before he sent it to anyone else, which she took as a compliment and also as a kind of thanks, though he did not say so.
There were four more events over the following five weeks, two of which held up to scrutiny and two of which did not. It was not the abundance of data that Samuel's thesis had imagined. It was, however, enough to establish that Venus was seismically active at a level that several models had predicted and several others had not, and to place a first, rough constraint on the structure of the planet's crust. It was the first such measurement ever made on any world other than Earth, the Moon and Mars, and it would be, Priya thought, the foundation of his career, if anything was.
*
Arden fell silent on the hundred and fourth day. The battery's voltage had been declining faster than predicted for some weeks, and the engineering team believed, though they could not prove it, that one of its cells had developed an internal short. The last relay pass brought down a weak carrier, then nothing. There was no drama about it. The lander had outlived its predecessors by a factor of more than a thousand, and nobody in the operations room was inclined to complain that it had not outlived them by a factor of twelve hundred.
In the weeks that followed, as the teams turned from operations to analysis, Priya found herself thinking less about the lander, which had done everything she had built it to do, than about the allocations. She had made perhaps a hundred of them over the course of the mission, large and small, and each had sent some portion of the planet's information to one group of people and withheld it from another. She was aware that the record of the mission would show what had been learned and would not show what had not been learned because of her choices. The thinned weather averages, the chemical measurements not made, the dust devil lost: these would leave no trace, and there would be no one to say whether their absence had been worth what was gained in their place.
It is the condition of most useful work that its costs are invisible to those who benefit from it, and its benefits invisible to those who pay. Priya would not, in the ordinary course of things, have expected anyone to understand what she had done; and she was surprised, a year later, at a conference in Houston, when Isabelle Garnier sought her out between sessions and told her, without preamble, that the two-hourly record had turned out to be sufficient for everything her team had wanted to do with it, and that the high-rate wind data captured during the seismic triggers had, quite unexpectedly, been the most useful part of the meteorological dataset, because it showed the structure of the surface winds at a resolution nobody had thought to ask for. She said this with the same correctness she had shown in her email, and then she went back to her session, and Priya did not see her again for some years.
*
Samuel's paper was published eighteen months after landing, in a good journal, with Priya as the last of eleven authors. He was offered a faculty position the following spring. He wrote to tell her, at a more reasonable hour than was his habit, and said that he did not think he would have got it without the reallocation, and that he had been more frightened during those weeks than he had let anyone see. She replied that she had guessed, and that she was glad, and that she hoped he would remember, when he came to make allocations of his own, that the people on the other side of them were frightened too.
It is not given to many of us to do anything that history will remember; and of those who are so favoured, a good proportion owe their fortune to the quiet decisions of others whom history will not remember at all. The plains of Lavinia are as hot as they ever were, and Arden sits on them still, a small box of silicon carbide slowly corroding in the carbon dioxide, its battery dead, its radio silent. It is a monument to many things. It is also, though no plaque will say so, a monument to a hundred mornings on which a woman in Altadena opened a spreadsheet and decided, as fairly as she could manage, whose turn it was to be disappointed.