Curiosity

by Rod Pyle

  The room was probably fifty by sixty feet, all in government-mandated gray and white. Not World War II–issue, but definitely decorated with a limited budget in hand. Little gets wasted at NASA, and you can feel it. Even JPL has a tone of spare efficiency—whatever it takes to do the job and little more. The wonders of the universe are yours, but don't expect any extravagance. In the center of the room was a cluster of long office tables, wired up with teleconferencing gear and hookups for laptops. Science in the twenty-first century is conducted via laptops, and to my inner satisfaction the vast majority of these were Apple MacBooks. The perimeter of the room was lined with more working spaces, each adorned with a PC tower (remember those?), most running on Unix. Efficient, if a bit Spartan.

  Webster and I settled into a corner, and while I prepared to take notes he nodded hello to a few people. He has been at JPL for over a decade and probably knows more about what is happening on Mars at any given time than any other civilian, yet there is a humility about him that impresses. He's a smart and capable person, about my age but (in stark contrast) tall, lean, and trim. There is little pretension in him—he came to JPL as a newspaperman years ago (when such jobs were far more common than they are today) and stayed on to run public relations for the Mars exploration program. He clearly has vast amounts of respect for the people he works with and lets it be known how much he enjoys the opportunity to sit in on these sessions. Where some in such positions might have been checking e-mail on their phones or gazing out the window, Guy Webster attended to the proceedings with rapt interest, as did I. But he's done it a heck of a lot more times than I have, and I admired his enduring enthusiasm. I've also noticed that a few months later he is likely to remember details from the meeting that have fled my mind. That's the kind of person you want in this kind of job.

  A few of the participants for this morning's meeting were already in place, quietly catching up on the dozens—sometimes hundreds—of e-mails that have stacked up overnight. It's a steady stream of mission-related correspondence. They are a busy bunch.

  As the rest of them drift in ahead of the start time, I begin to get a bit uncomfortable. Why? Oh, I see—I am in a room full of children. At fifty-seven, I am probably the oldest person in the room (Webster may have a year or two on me—if so, it is much appreciated). The Mars exploration program is run by people who look like grad students with a few premature gray hairs here and there. I know that a couple of them are in their forties, but you'd not guess it if you didn't know—such is the preservative power of having passion for your work.

  We were scheduled to attend two meetings today: the first would be a broad daily-report sort of affair, and a couple of the high-level people sit in for that. Ashwin Vasavada, one of two deputy project scientists and a direct report to Grotzinger, was there. The meetings were to be chaired by Bethany Ehlmann, who is in her early thirties and already is an assistant professor in planetary sciences at Caltech as well as a participating scientist on MSL. She radiates unassuming capability.

  Daniel Gaines sits to the side. He is the a senior member of JPL's artificial-intelligence group, and today serves as a tactical uplink lead, overseeing what will be sent up to the rover at the end of the current shift, instructing it what to do tomorrow. All of these people have titles, but few constrain themselves to that area. There is just too much to do, too much to learn.

  Dan Limonadi sits next to Ehlmann. He's in his early forties, came to the United States from Iran as a child, and quickly realized that his fate rested in space science. He is a lead engineer for surface sampling and a specimen of a man—tall and rugged looking (where do they get these people? It looks like an H&M commercial in here). I've spoken to him before, and he is also a pillar of the community outside JPL—he performs all sorts of community good deeds, works on the local search-and-rescue team looking for lost and injured hikers in the local foothills, and coaches some sport or another. I rarely found time to even attend my kid's soccer games, much less coach them. Sheesh. Limonadi is busily preparing for his part of the meeting.

  A couple of rover drivers, a deceptively simple term for people who need to understand every detail of how the rover functions while crossing terrain, sit nearby. One is Paolo Bellutta, who will begin the discussion about driving goals for the upcoming sol. Bellutta also distinguished himself by developing software to help in the selection process of a landing site by assessing drivability. The other is Mark Maimone, whom you may have seen on camera if you keep up with JPL's videotaped rover updates. He is one of the few people here who is not rail thin, and I silently thank him for that. But he has a doctorate in robotic sciences and is better looking than me. Oh well.

  Beth Dewell is a tactical uplink lead, and her work dovetails with that of Gaines. She looks young and sounds smart, just as they all do.

  Youthful though they are, most of these people have experience dating back to the MER rovers Spirit and Opportunity, and some to Pathfinder. They are the senior (in experience, not age) faces of Mars surface exploration post-Viking. And many have doctorates, and if not, technical masters’ degrees. That's a lot of education in one small room.

  The meeting is called to order—it is not a stuffy affair but rather is collegial and delightfully informal. Nonetheless, it is efficient and the business gets done. The official name for this assemblage is the Science Operations Working Group, or SOWG. It is not the best acronym at NASA, conjuring up images of something that might end up on my breakfast plate (I'm not a vegetarian). I prefer to think of them as Curiosity's caretakers, minders, and teachers. The machine may be semiautonomous, but a lot of smart people are behind what it does each day—excuse me, each sol. Many of them are here today.

  The meeting includes people who are phoning in from remote locations, and it begins with a discussion about Curiosity's current uplink and downlink status, power levels, and other immediate concerns. Nothing jumps out here, and it sounds as if all is well. Of course, I am a tactical-planning-meeting newbie, but Gaines said as much, and everyone seemed pleased, so I take that as a good omen.

  Next they talk over the terrain surrounding the rover. Remember that Curiosity can drive autonomously, and do so farther and somewhat faster than its predecessors could. So knowing what is nearby and in its path is pretty important. First it makes everyone's job easier, including the rover's. Second, it avoids any nasty surprises. One of the drivers notes that here is a precipice up ahead. It's impossible from this angle, either from the Mastcam or from orbit, to really glean quite how far it drops, so they are exercising caution. It does not appear to be anything huge, but it does not take much of a drop to cause problems. A hard right-hand turn would provide a safer, more predictable detour, but adds dozens of feet to the drive to Mount Sharp. At these speeds, this kind of distance matters—Curiosity's top speed is 1.5 inches per second, or about 450 feet per hour. That's less than a quarter mile per hour. And at most times it will not be driving near top speed, so this gives you an idea of why distance is important.

  But safety matters more than speed. It's at times like this that you realize, despite the challenges and hardships it would entail, just how nice it would be to have a human being up there on Mars. An astronaut driving NASA's lunar rover from the 1970s could drive over far rougher terrain at about 8 mph, topping 11 mph when pushed. It's that human Mark One Cranium computer at work. And, of course, humans have an innate ability to fix things that might go wrong on the surface so far away. You quickly realize just how vulnerable robots on distant worlds are—mishaps can quickly be fatal. Patience and planning are the watchwords.

  The group gathers around a pair of monitors. One side shows a 3D model of the terrain surrounding the rover, including a blank patch where the precipice that has them concerned is located—no data. The other shows a photomap of the same terrain.

  There is another issue to consider. The area they are crossing is hard ground, like a packed, cemented lakebed. Normally this would be great news because hardpan surfaces allow for faste
r driving—nary a sand dune or trap in sight. But—and it's a major but—there is a problem with this hard surface. There is a bunch of small rocks all over the area. Normally these rocks, maybe the size of baseballs, would be a nonissue. But they are sharp and potentially dangerous to the rover. Apparently, millions of years of wind-scouring resulted in a bunch of jagged points sticking out of the area like a parking lot nail strip. You know, the ones that are to prevent you from going out the wrong driveway, and if you do, rip your tires to shreds.

  Sometime earlier, the techs had been performing the usual “let's look her over” picture sequences with the MAHLI arm-mounted camera pointed back at the rover. Something was not quite right—there were extra holes in the wheels. The wheels are supposed to have some holes. There are a series of them, which are used to track wheel revolutions via their imprints in the soil—there are slots and circles machined into them that spell out JPL in Morse code. The lab had wanted to just etch letters, but NASA nixed that, either out of design concerns or possibly institutional envy. So Morse code it was.

  But those are cleanly machined holes. Now there were extra dents, holes, and even tears in all the wheels. Recall that these wheels are about the diameter of small beer kegs. The rover they support weighs a ton. If they were made thick enough to resist all damage (each was machined out of a solid block of aluminum), they would weigh far too much. So instead, they have the strong areas where the cleats or ridges are, but the connecting metal around the diameter, between those cleats, is pretty thin.

  The wheels can take a lot of abuse before failing, but nobody wants to even come close to that point. It's early in the mission, and they have a long way to go. It's just another thing to worry about, another obstacle to quick progress. They will have to drive slower to avoid the worst rocks and minimize the damage. Mount Sharp just got a little farther away.

  As this debate is concluded, I notice that one of the guys on my left is typing commands into a computer. So what, you might ask? It got my attention because the information on the screen was just lines of green text on a black background—probably in Unix—and looked just like my first computer, an IBM XT, from 1984. Déjà vu, flashback, my goodness I am old. Sigh.

  Vasavada caps the debate by saying that the choice of driving route is not a decision for the science folk, it's one for the strategic planners and drivers. The longer route gets the vote, even though going straight would save several sols. Safety first.

  On to the next topic. The rover's state of electrical charge is discussed. Curiosity's nuclear power plant is like a trickle charger on a car battery. Good for supplying a continuous few of a little bit of power, but not nearly enough for the real-time operation of a machine of this size and complexity. The nuclear fuel charges the batteries, and the rover runs off them. The state of charge, essentially how much juice is in those batteries, is affected by the amount of work the rover is doing, what instruments are on, how fast and how far they drive, ambient temperatures, heater use, and more. So it's a constant concern. Right now, it's reported that it's a bit lower than anticipated but still in an acceptable range. They agree to keep an eye on it, next topic.

  A few more technical points are discussed, and one thing I get out of it is that the current light-time between Mars and Earth (the one-way time that a radioed message takes) is ten minutes, twenty-seven seconds. Interesting.

  The first meeting adjourns and Vasavada and a couple others depart.

  The next meeting is for more immediate day-to-day operators and the science team representatives. They need to plan more minute details. By golly, I thought that the previous meeting was about the minute details. Nope.

  Ehlmann conducts a roll call as the people around the table identify themselves for the benefit of those attending online. The meeting kicks off with Ralph Milliken, who is attending remotely and is an assistant professor from Brown University. He looks to be about thirty from his photo; I suspect that he must have received his doctorate when he was nine….

  I've become age-obsessed. Moving on.

  He refers the team to a topographical map of the area, then discusses the Mastcam test they are trying to complete. Also, there is an outcrop nearby that would be great for the geologists (he's one of them) but probably not so good for the drivers and the rover's wheels. They collectively look at another image of a punctured wheel, and that discussion begins anew.

  Then plans for the next few days are discussed. These involve instruments and the tasks needed to operate them at each juncture. Certain tasks require the machine to be at rest, others don't. For tomorrow they plan to do what is called “untargeted remote sensing,” which means in effect telling the rover to keep her eyes open and record what she sees as we drive (everyone calls Curiosity “her”; guess I should start as well). Then on the next two working days, Friday and Monday, they plan to do “contact science”—the robotic arm will reach out and touch a rock or soil patch—which obviously requires the rover to be still. During the intervening weekend, there will be more contact science, and the team discusses the high- and low-priority targets.

  Once Curiosity has these instructions for a specific sol, conducting the science becomes an autonomous process. Of course, many procedures are monitored from the ground, which is especially important for the people watching the arm when doing contact science. You're placing the arm close to something solid and unyielding, usually a rock of some kind, and you do not want to either bang the arm into the rock or have the rover shift or slip while you are near the rock. You also don't want the arm to hit or scrape any part of the rover while it is being positioned. These sound like minor concerns, but with a one-ton machine sitting behind the arm, any unintended pushback or hard bumping could be disastrous. The onboard software is constantly vigilant and programmed to avoid most potentially dangerous situations, but it cannot know everything. It needs updates for new situations.

  The following Tuesday they will use the MAHLI camera to look over the wheels again. Today they noted one new pinhole and small tear on one of the wheels, but nothing showstopping. Just more slightly above routine wear. The wheels can take a lot of abuse, as the surface area in each is so much more than this level of damage can compromise.

  A set of rover instrument conditions is surveyed in preparation for another cold Martian night:

  Mobility heating—keeping the drive motors happy.

  Arm heating—keeping the joints on the robotic arm limber.

  DAN passive—the Dynamic Albedo of Neutrons instrument is not working in active mode (i.e., not actively shooting neutrons into the ground), but is passively reading neutron activity in the soil.

  Mastcam heating—the cameras on the camera mast will be kept warm.

  MAHLI heating—the microscopic imager on the arm turret will be heated.

  Navcam heating—ditto.

  Ehlmann, who is also chairing this meeting, performs the time-honored tradition of the go/no-go poll of the principals in the room. This process was termed “going around the horn” by Gene Kranz, a flight director in the Apollo lunar landing days. The term was not applied here, but the process was the same. Everyone said “Go.”

  There is then a detailed discussion, with some dense graphics up on the big screen overhead, of various technical issues. The last data downlink, which gives them, among other things, the information needed to plan the next day's drive, was not as much as expected. Some kind of bandwidth bottleneck. They will plan another downlinking session with the Mars Reconnaissance Orbiter, one of the two orbiting data-relay conduits available.

  As mentioned, they will be handing over the rover to the next shift with a bit less power reserve than they would like. They will shorten the drive tomorrow to compensate and move some of the post-drive imaging targets to a later time. Lower than expected temperatures are affecting the batteries, draining them faster than is optimal. But it is all well within acceptable limits.

  James Biehl, who joined this meeting after the first one adjourned, talks a
bout the upcoming terrain and driving decisions in more detail. He joined JPL in 2010, and from the looks of him, that was fresh out of grad school. He appears to be a smart and intense kid. He announces that they plan to instruct the rover to drive twenty-six meters tomorrow, or about eighty-five feet. This will be accomplished in a little under an hour at the end of the sol, allowing the rover to perform as much science as possible while it sits in the Martian sunlight. They will snap an image of the wheels frequently to make sure that the drive is not causing any more damage than is absolutely necessary.

  A plan for activity with CheMin is discussed, and the restricted data flow comes up again. It should not stop the instrument's activity, but it is on everyone's mind. They look over all the data modules that need to come down from the rover, some technical, some scientific. Priorities are assigned: critical 1, critical 2, etcetera. This way they can get what they really need to move forward and save some of it, mostly science and rover-status data, for a later data-relay pass. Curiosity's computer is, of course, capable of storing information for later uplinking, and the orbiters can do the same if need be.

  The meeting begins to feel like a technological ballet. This is the detailed working session, and everything is presented, discussed, analyzed, prioritized, and positioned in an exquisite fashion. It is an amazing synchronization between departments, between instruments on the rover, and between the rover, the orbiters, and the ground stations in the Deep Space Tracking Network. Even the resolution of the images to be downlinked is discussed, as lower resolution leaves more bandwidth for other information. Higher-resolution images can be stored for later if they are later deemed important.

  As regards data flow, they discuss windows of availability. The Mars Odyssey pass, later that evening, will allow for seventy-four megabits of information. That's only about ten megabytes, which is not a lot of bandwidth. The Mars Reconnaissance Orbiter will provide more, as it has a larger (almost ten-foot) radio dish and what is called an “adaptive data rate.” It can adjust to the needs imposed upon it. Incidentally, noted on the complex spreadsheet-style grid filling the big screen is leftover data from a week ago—there are science people still waiting for results and information at least five days old. They will download it when they can. None of it is critical, but I can visualize some foot-tapping going on wherever some of these scientists who are joining us online are based. Science takes patience, but it's easier to wait when data is being gathered and crunched than when that data are sitting in storage, waiting to be transmitted. But everyone involved is reasonable today.