Thursday, August 25, 2011

A Week In the Life, bSols 2694-2696: Like A Big Camera Saw


If the Microscopic Imager (MI) was a saw, we'd have two Tisdale-2's.

If the amount of action in MER town were anything to go by, it seemed like it was a few weeks since my last tactical shift as TAP/SIE. It was more like 10 days — long enough to get me itching for some more. Lucky for me, I just came off a slew of 4 planning shifts in a row: Friday, Monday, Tuesday, Wednesday. They say you get a much better perspective as a tactical engineer when you do consecutive plans, and boy are they right.

My last shift was for sols 2683 and 2864, when we were planning a drive to get the first good shots of the inside of Odyssey crater. Since then, we've parked in the ejecta field from this small crater:



We imaged around a bit (notice Tim Parker's overlays of the rover images!), taking our foot off the pedal, and selected a few beautiful targets. The one of most importance? Tisdale 2:



With its flattened and gold-dipped top — suspiciously looking like an aircraft carrier deck — and its layered "breccia" appearance (ask a geologist, I'm just here for the lulz), the SOWG picked it as the target as the next victim of the IDD. The failed — errrrr, slightly off nominal — drives left us in the situation where the RPs needed to really do their homework. Indeed, Paolo and Kevin pulled off one hell of a drive on sol 2692. It was complete with a nice turn in place and a careful approach to get the center of the rover within about 1.2 meters of the edge of Tisdale 2:

(courtesy mhoward on UMSF!)

They nailed it. 

We walked in on Monday to plan Sol 2694 with a fresh sequence crew and a ready science team. Looking ahead, I saw two weeks of Nominal and one week of transitioning to Restricted planning via Slide planning sols. That means heads down, power through it, we've got some IDD'ing to do. The IDD work volume appeared right on our screens, constrained by the busted shoulder joint to be a straight plane. We'd be pushing the boundaries this week.

The idea of this week's IDD campaign on Tisdale-2 was to effectively make a run of MI, MB, and APXS measurements up and down that work plane, from top to bottom of the edge of the rock. Looking at multi-spectral and contrast-stretched images, we could see all sorts of interesting features. The SOWG teamed up with the RPs to plan a series of 4 targets, appropriately named "A/B/C/D," for the usual suite of IDD work:

1) Touch the target with the MB to get an accurate idea of where the surface is
2) Take some MIs
3) Rinse and repeat

On the last target, we drop the APXS and let it blast (lightly!) the surface with its radioactive Curium-244. Our Navcam PUL filled out a missing part of the surface and sky to the rover's left, filling out a nice mosaic for targeting in future sols:


On Tuesday, our downlink was a little thin — we could barely tell what had happened, only that we had successfully placed the APXS (meaning the whole MB/MI/APXS sequence executed nominally). Very few images had come down. Diagnosing, we found out we were just victims of bad luck with poor weather at the DSN station; looking into the future, we see more scant downlinks and few opportunities to make up the lag.

(Every sol, the LTP report contains a recommendation and a qualifier — "thisol's collection limit is 150 megabits, but it's a soft requirement" — that accounts for both the desire to collect new data, the downlink capability for that sol, and the long term picture of what our downlink schedule looks like. Every sol from 2964 through 2996, the LTP recommended the usual number — call it X — and we tended to more than double it. We understood the consequences. I always liken the collection of data and amount of stuff left over after each sol's downlink to eating and weight gain: Just as you'll see a few days' lag from over-eating and weight gain, you'll see a few days' [sols'] lag from data collection to flash volume left over.)

Sol 2695 was more of the same: Pick a few new spots, and do the MB-MI-APXS suite of measurements on each. These were a little more challenging: Ashley, verified RP ("IDD pilot") pointed out that one of the spots was extremely close to one of the other spots from sol 2694; there was concern about having too much soil on the lower spot of the rock to make the APXS useful; could we reach the top of Tisdale-2?; etc. In the end, we did another top-to-bottom approach to moving between targets. The slew of targets on the screen looked like a dartboard, especially when you layered in the targets from sol 2694:



Don't be fooled by the perspective: There's a little bit of yaw that you can't "feel" in this image, and remember that the wrist of the IDD has another degree of freedom, meaning that we can actually reach out of that constrained work plane by a limited amount. With this set of targets, you get the feeling that we're cutting this puppy in two, trying to figure it out, trying to tell a story. 

Already, the scientists are buzzing about the composition of Tisdale-2. What's unique about this is the tactical role of science on MER: with rovers, many of your decision points are determined by the science you collect. Say the APXS comes down — boom bam, quick analysis, yeah it looks good, we can move on — suddenly you've got a wealth of power at your hands. You can stop and breathe if you need to. With orbiters, you've usually got only one shot to get the geometry right, and you won't be coming back any time soon. Moreover, the science you collect certainly doesn't feed into decisions made tomorrow. Or even next week. Takes a little bit longer. With rovers, you get a closed loop feedback cycle from your science team. Like this:


Reader, Tisdale-2. Tisdale-2, reader.

Then came sol 2696 ("Where did the week go?" asked Mission Manager Scott) — more of the same. A few new targets, a few more MIs. Same deal. Only… these were complicated. In the pilot seat for the RPs were the illustrious Scott, Vandi, and Julie. The targets that the SOWG wanted were very difficult to maneuver to. There were a few "wrist flips," IDD joint moves, and off-plane target goals. When queried for a time when they'd be ready for APAM, the RPs paused: "… 1pm."

Ok, cool, perhaps we're in it for the long haul. Our uplink was a little earlier that night than we wanted, but it was manageable. But I looked at my schedule from Monday and Tuesday (planning sols 2694 2965, respectively) and noticed a slip every day in the APAM time: almost an hour later every day. Historically, APAM tends to be a good measure of the length of the planning day, if we ignore those pesky anomalies when we're there until 8pm. When I peek back at my schedule over the last year, a 1pm APAM tends to mean a long day. 

Then again, there was a distinguishing quality of those longer planning days: They were caused by curveballs in the planning process. A script breaks here, an integration goes awry there, someone spills coffee on the keyboard. On this occasion, sol 2696, the lateness of APAM was mandated, controlled, intended. This bodes well because it is an act of foresight, and intuitive measure of how crappy things might get. We reap all sorts of benefits from this (though not always), the biggest two of which, I think, are:

1) Extra intended time means extra preparedness for the next walkthrough.
1) Everybody has the mindset that we're up to our neck in this sequence, and we better bring our A-game.

When a delay is unintended, we are off the balls of our feet, off axis, expecting the worse, freaking out a little. Attitudes towards computers grow increasing hostile. But when we mean for something to be late, we're on the ball.

I wrote that thought down around 10:30am that morning, not thinking I'd be so clairvoyant: Indeed, APAM was a quick 15 minutes; the next walkthrough was a mere 20 minutes; and we had the sequence ready for delivery before yesterday's plan. Being prepared actually bought time back. Everybody was on the same page at every meeting. Snip snap done. Like that.

—————

Oh! Right! This week's albums of the day:

Sol 2695: Surfin' USA — The Beach Boys**

*Actually, this is from a different album. Meh.
**This was Brenda's request. I only abided because she's nice to me and she was shadowing me on TAP/SIE duty on Tuesday. I hate the Beach Boys. No video for you.

Saturday, August 20, 2011

From 11's to 10's

Big fat analogy coming. Closing this window is for quitters. And I know you ain't a quitter.

—————————

The other day, I was talking to a friend that I hadn't spoken to in a while. She asked a pointed question:

"What's new?"

My answer was less pointed: 

"Work's getting busier. Ankle still on the mend. Oh, and I switched from 11's to 10's on my guitar. This is a momentous occasion."

That last bit was a little cryptic. Not being of the musical type, much less of the guitar type, she didn't have a proper response. Indeed, it was a momentous occasion when I made the decision to go from 11's to 10's, but she had no idea why it was so momentous. 

A little background: "11's" or "10's" (or "9's" or "12's" or…) refer to the size of the strings on a guitar in units of one-thousandths of an inch (i.e., "11's" = 0.011 inches). Saying the one number like "11's" or "10's" on its own doesn't tell you full the story because there are, of course, five other strings for a standard 6-string guitar whose sizes haven't been named. Saying "11's" or "10's" refers only to the first string. However, it still gives you a good approximation of a guitarist's preference. Usually, "11's" means "medium size", and your strings range in size somewhat nonlinearly up to 0.048 inches; we say "11 to 48's" for shorthand. For "10's", or "lights," the usual range is from 0.010 to 0.046 — "10 to 46's". (I say "usually" as there many exceptions to the standard.)

Get it? Got it? Good. Bear with me.

10's (ahem, 10-46's) are the norm for guitarists. I'd bet you a good sum of money that when you walk into a Guitar Center and start playing a guitar, the strings on it are 10's. On the other hand, 11's are considered heavy, even though they are called "medium" as if to suggest that they are "average" in size. 11's are popular with rhythm guitarists, as they carry a little bit more oomph and sustain. As a consequence of their larger size, they require a little more finger strength to play. You can, however, "hide" under the heaps of additional tone that you're getting with them.

When I started playing the guitar in the 9th grade, I was playing with the standard 10's. In my senior year, I managed to build a rather unhealthy obsession with Stevie Ray Vaughan. He continues to be one of the most prominent inspirations in my guitar playing. What's important to know about SRV, other than he was an unequivocal god on the guitar, is that he played super heavy strings: 13's to 62's.

That's heavy. Really heavy. 

It's damn near impossible to play a guitar with 13's. Funny stories of SRV on the road tell us that he super-glued his fingertips because he played so hard on such heavy strings. (To be fair, he also had to tune the strings down a bit to loosen the tension.) If you catch a video of him in just the right light, you'll see stuff flying off the neck of his guitar: it's either skin or super glue. 

Awesome.

SRV used his guitar as a percussive instrument. It is difficult to make this sound good with light strings because light strings have less bottom-end, less bass, less tone, and when hit viciously hard they break or sound like brassy junk. (You can hide the crappy tone under a Marshall head and cab, making it sound like NASA built your rig.) Because his strings were so heavy, his percussive playing added depths and layers and tones heretofore unknown. It just sounded big. He could also play pretty damned fast, but let's remember that Texas blues ain't about that. Texas blues are about big, sloppy, heavey, percussive, toned, steroidal lead guitar, about blues taken up to 11.

Figure 1: But these all go to 11.

"Sloppy" is the appropriate term here because, like I said, you can "hide" under the percussiveness. You can have a little bit of error in your playing because you aren't trying to play at lightning speed and you certainly don't give a damn about missing a note here and there. You're earning your notes, but in these big huge batches, not one-by-one. It's a macroscopic accumulation only visible with a wide lens. The nitty-gritty don't matter. It don't matter one bit. (Ya'll.)

During my SRV obsession some seven years ago, I took my beat-up and used American Stratocaster and beefed her up with 12's — 12 to 56's, to be precise. The 13's were too much. After I had saved enough money, I bought myself a Gibson Les Paul:

Figure 2: Me and Miss Paul

The Les Paul is not a guitar that's used to having anything more than 10's on it. Jimmy Page of Led Zeppelin used a Les Paul, and his only advice to young guitarists was, "Use light strings." My SRV obsession still hung over my head as I made the purchase, so I was compelled to make a compromise. I settled on 11's — 11 to 52's, actually, not 11 to 48's — giving me that classic "regular-top-heavy-bottom" sound. Lots of boom on the backside with enough shaved off the lighter strings to give me some breathing room when I'm playing those face-melters high on the neck.

Seven years later, my SRV obsession has subsided. With that obsession has also gone that necessity to make a compromise between sounding like SRV on a Les Paul (incidentally, an impossible task for a Les Paul, much less for anyone besides SRV, who was not human) and having some fine-scale noodling capabilities on the high end. I settled on 10's to get some lightning speed, something of which I've had only inklings in the past. Pure speed with some gas behind it is hard to come by, and 11's sure ain't helping for a mere mortal like me.

I removed my skin-tainted and dirty set of 11-52's this morning so I could trap on the first set of 10's that my Les Paul has ever seen (well, since it's inception at the guitar shop in 2004). The transition to 10's has been tough: my vibrato is all out of whack; I'm over-bending; my percussive sloppiness, emulated from SRV and Joe Bonamassa, is coming to light and the veil has been lifted on my mistakes; I'm less nimble than I thought; I'm hitting and gripping way to hard, being used to having to work harder to play each note. It'll take a week or so to get back on track.

Get it? Got it? Good. Bear with me.

—————————————

This past week in the world of MER-B, we had a few problems with our drives. Per the status update on the MER website:

On Sols 2683 and 2685 (Aug. 11 and 13, 2011), the rover performed a pair of drives to position herself for a close approach to the rock target. On Sol 2688 (Aug. 16, 2011), the planned approach drive stopped early because the rover's visual odometry could not measure progress accurately due to a lack of visual features in the camera field of view.

Our tactical meetings this week were also a little more… strained. These short drives require a much higher level of detail in the planning and we're mixing in some complex observations. The geometry is awkward, the terrain is deceptive, and the coffee isn't flowing at a high enough volumetric rate. Everyone from the RPs to the TUL to the TAP/SIE is feeling it. There are curveballs at every corner, waiting to be swung at — and missed.

That said, these challenges are well under our umbrella of capabilities. To call them "growing pains" would be inaccurate, as most of the tactical team has done this all before, complex drives and elaborate system-level sequencing alike. It would be more faithful to the current situation to call it "morning drowsiness." As I've explained before, I haven't seen such involved sequencing before because I joined the MER project when we were in the middle of nowhere.

When we were doing those big drives, there was a little bit more margin for error. The terrain wasn't dangerous and we were, most of the time, building plans around drives. Images were taken tens, if not hundreds, of meters apart. A botched plan meant a missed day over a timescale of months. It was pedal to the metal. It was go big or go home. It was macro. We could have a little bit of slop — within obvious limits — and we could "hide" some detailed, fine-scale errors because they didn't really matter for the Meridiani dune fields. We needed accuracy, not precision; macro, not micro.

Now, we're building drives around plans, the original focus of the mission. We didn't blast a couple rovers to Mars to drive them, we blasted them there to let them do science. Driving is a capability, a means and not an end. This is a crucial concept for the team to have in their head. I've never experienced it, so it's new to me — growing pains — but the team has experienced it, and it's not new to them — morning drowsiness. We're still waking up to the fact that, oh crap, we're at Cape York and it's time to do some science.

Figure 3: Science is required in this hizz-ouse (courtesy Mike the Mission Manager)

Brush the dust off the manuals. Have another cup of coffee. Limber up those engineering muscles. We've got some science to do.

The analogy is thus…

For the last several years, Oppy's been in cruise control. The sequencing detail has been at the fine scale only where it needed to be. Things were macro. Plans were focused on the large-scale picture of "get some mileage, then catch your breath." A botched plan can mean several missed days over a timescale of a few days. It was about accumulation of tracks, appreciation of distance, at the risk of system wear-and-tear. We were percussive, hiding the micro-level details in the noise of huge drives and well-pointed imagery. We were playing with 11's.

Now, we're at the Cape. We're about precision, not accuracy. We're about the micro, not the macro. We're about short, expertly-planned drives and precise, expertly-planned imaging. We're about near-term plans, not mileage. There is no "Pink Path" to which we must abide. We are now playing with 10's. And the adjustment is hitting us. It's taken about a week or so to get back on track.

We got lucky in two departments: 

1) The adjustment hit us while we were in Restricted planning sols. We were driving every other sol at most and planning every other day at most. There was time to catch our breath and think twice about our vibratos and bends and finger strength drives and imaging and systems engineering. 
2) The adjustment didn't really hit us that hard. We aren't frustrated or suffering. Just acclimating to a previously-known norm. We didn't put the rover in danger and lost, maybe, a sol or two due to the botched drives.

Monday, we hit Nominal planning. And it's not pedal to the metal, it's science to the… metal. (?)

——

For those that are interested: The "No Shave Noachian" chinstrap has made some progress. Composite photography and growth rate analysis suggest that the current appearance is as such (accuracy not guaranteed):


Tuesday, August 16, 2011

On Coffee Cups and Restricted Planning


Today is a momentous day, for I have purchased a new coffee mug. For almost a year, I have been sporting the standard JPL coffee mug. It’s blue and generic. Also, a piece of junk. Also, I hate it. It waves the banner of inadequacy, thinking that its deficiencies are something to be proud of. Speaking of its deficiencies, let's focus on its structural integrity – or lack thereof:



A drop in the sink and crackkkkkk. Son of a biscuit.

Since the Mug Drop Incident of 2011, I have been going with an old standby:



It’s a remnant from the LASP days. Sufficiently stained from lack of washing, it’s like cast iron skillet of coffee mugs: flavor baked right in. Eat your heart out, Paula Dean.

Of course, the standard open-ended coffee cup provides no longitudinal spill protection. If the local gravity field suddenly flips upside, what in the crap am I supposed to do? What I needed was a new portable mug. The JPL store didn’t have any good ones until recently. This morning, though, I poked my head into the store… and behold!



Its major features: 
1) Unimpeded Coffee Delivery Orifice (CDO)
2) High-rate Air Exchanger (AE)
3) Aerodynamic design for hypersonic reentry
4) Thermal protection for long-duration coding binges
5) Easy-pop lid (EPL) for tactical refill operations
6) Thumb indent for maximum grippage
7) The aura and presence of a true coffee mug, making me look seasoned and important

Right. Now. On to a real topic.

——————

Determining when a given planning day is “restricted” – roughly meaning we don’t know the full the state of the rover and more than a full sol is still left to execute before the sol(s) in question today execute – is a non-trivial process. It’s a function of many things, including but not limited to: 

-Allocated time with the DSN
-Direct to Earth (DTE) and Direct From Earth (DFE) communication window times
-Earth-Mars Geometry
-Earth planning day constraints
-Per-sol downlink latency
-Presence of dragons
-Coffee left in the cup before refill required

Some of those things we can't control. Some of them we can. We've got a guy that puts all of this information in a giant spreadsheet and works some magic. With a wave of the hands and a refill of the coffee (I’m making the leap of faith that he does drink coffee, which is, of course, a reasonable assumption, as you can’t be human and not drink coffee), the spreadsheet tells us which days (sols) are nominal, which are restricted, and which are ‘tweeners – slide and tight.

We spend up to three weeks in restricted planning mode; we usually get about two weeks of nominal, and a maybe week of tight/slide bookending this. The cycle is never the same, but it’s always predictable. 

There are several curious aspects of restricted planning. First and foremost, there is a change of gears throughout the project. There are, at most, 3 tactical planning days between Monday and Friday during a restricted week. This means less tactical shifts to go around. Even if you are, say, scheduled for up to half of your working week to MER, you probably won’t meet that in restricted weeks.

For multi-mission goonies like me, this opens some other doors. My mindset goes to another tier; not a lower tier, not a higher tier… just a different tier. I always say that, “I can go get real work done now.” I bury this statement in a few layers (at least seven,) of irony to make sure nobody takes it too seriously. 

By way of example, for a few weeks I’ve been involved in what we’re calling an “Agile Science Team” study to investigate fast-reaction-time mission planning time for spacecraft like Dawn and Rosetta. It’s cool, because it’s all orbital mechanics and attitude dynamics, it’s all fields-of-view and line-of-sight, it’s all dynamic events and comet plumes, and it’s all high-level abstractions. 

The lack of scheduled tactical time with MER means I get to focus on side tasks like this one. For this task in particular, this free time bit me in the ass – precisely because, as it turns out, it’s very difficult to say anything truly meaningful about a spacecraft trajectory. It's just a path in space, and it means nothing on its own. The abundant free time allows me to dwell on this task almost too much. Four hours of staring at plots later, I’ve got nothing to say except a few hundred words in a Word document.

Trying to get this same kind of thing done while buried in MER tactical work is more of an adrenaline rush. Remember, when on MER tactical duty, all other work takes second priority. However, you can find those nooks and crannies in the day to answer those half-dozen questions and email threads sitting in your inbox. Your answers become quick, to the point, lacking in capitalization, devoid of recognizable grammatical structure. The chances of this adrenaline-fueled answer being useful is about 50-50, a veritable shoot-first-ask-questions-later shotgun approach to the job. Sometimes you look like a total badass. Sometimes you just look stupid.

By way of another example, let's pretend there's a mini-crisis going down on another mission you work with while you're on MER tactical duty, planning nominal sols — day in, day out, day in, day out. A crisis like, I don't know… the ops guy spills coffee on his keyboard and radiates an accidental "backflip" command to his spacecraft. (Note to self: this command does not exist.) Something like that. An email races to my inbox, reading something like, "Anomaly… blah… coffee… blah… SOS." The adrenaline kicks in, just like in the other case, but you have to practice subduing it. You can't pack up and leave — MER comes first. You can't gut-react with shotgun emails — you'll only add rumor and crisis to what is likely already a rumor-filled crisis. 

HANDS. OFF. THE. KEYBOARD.

In restricted planning, when my MER shifts are more rare, I can zip on up to help the crisis. It's about balancing priorities. Your ability to look like an idiot notwithstanding, you're able to focus your focusy focused focus on another group of people.

Of course, we have the usual slew of MER team meetings that get scheduled every week regardless of the planning schedule. During nominal planning weeks, attendance at these meetings is low. A good number of folks are in the sequencing room on tactical duty; the rest are supporting downlink analyses and the like to keep the tactical wheel well greased. Restricted planning weeks come along, and as if somebody said, “Free beer at the IST meeting,” the room fills up.

This week, there was a special presentation by one of our summer interns, Will. He danced around some subjects, presenting some well-made charts with worthwhile commentary to back it up. A few months back, we had Paolo tell us about some traversability analyses he did for the MSL project a few months ago, which was used to help them select a landing site. (I thought Gale Crater looked best, too, based on Paolo’s work, although the decision hadn't been made by NASA; besides, work like his is only one piece of the pie.)

Every once in a while, the IST team chief will catch us off guard: "Ok, if there are no more round-table discussions… it's time to review flight rules." [collective sigh] To make up for it, Project Manager John Callas likes to do a lunch-time movie. Recently, it's been one of the episodes of "From the Earth to the Moon." Sweet.

Then… then… we can feel those nominal sols coming back up. Everyone knows when it's coming: Fridays, usually 3-sol planning days, become 2-sol planning days. Then, it's game time: limber up your IST muscles, because it's non-stop insanity for the next couple weeks. Then we can breathe again.

Saturday, August 13, 2011

I didn't know her, but she seemed pretty cool — a belated MER-A obituary


In honor of Opportunity's arrival at — errrr, drive that took her right past — Spirit Point, I give to you something I wrote for a personal blog a few months ago. I cleaned it up (Cursing? What's that? I certainly don't speak to my mother that way.) and fit it into the mold of this blog. I hope you enjoy it. 

And, at any rate, it'll explain the title of this blog. 

———————————

It was July 21 of last year when I had my first tactical shift for the Mars Exploration Rover flight project, rover B — aka MER-B, aka MER-1, aka SCID 253, aka Opportunity, aka Oppy. It was a training shadow shift, intimately known to veteran TAP/SIEs as, "Sit and watch until you're bleeding in places you shouldn't be." 

Months earlier, before even signing the contract for the job out at JPL, I had let my head get filled with the notion that rovers weren't anything worth my time. Somehow, I was "above" them. Or, something. I had trouble explaining it. It had something to do with the mechanical nature of them, which caused an immediate association with mechanical engineers, which caused an immediate association with the experience of seeing thirty percent of the folks in my aerospace engineering class drop the major in less in than three months for mechanical engineering because it was "easier." (It wasn't. I guess the joke was on them.) Or maybe it was the stigma of being on a flagship project. I hadn't met a single person on the MER project and yet I already had this idea that I wouldn't be able to identify with them. Why should I mess around with slow robots when I could mess around with spacecraft that were moving at a few kilometers per second, high overhead, responsible for more than 90% of their total downlink volume?

Two days previous to that July 21 shift, I set up a quick meet-and-greet with the Integrated Sequence Team (IST) Chief, Matt Keuneke. I stepped out of the elevator, two floors above my office in Building 264 at JPL, to find a huge pop-out picture of an artist's rendition of the rovers on the opposite wall. The words "MARS EXPLORATION ROVER 2003" were screaming across the top of the mural. There was a sense of gravity to this place.

I met Matt outside the restricted-access door. We walked to his office, just adjacent to a room full of computers — this one was suspiciously empty, cold, dark — where he gave me the run-down on the mission: Who he was ("IST team chief"), what I would be doing ("tactical activity planning and sequencing integration engineering for Oppy"), who my co-workers would be ("Well there's the TUL, MM, TDL, RP1/2, and PULs, for starters"), how much time I'd be doing MER tactical work ("More if I like you."), and some administrative nonsense.

Yeah, yeah, yeah Keuneke — get to the juicy stuff. I need convincing.

No such luck. I left his office still unsure of the whole MER thing. Was I pigeonholing myself? Was I heading into a deep dark realm from which there is no return? 

Are rovers actually… cool?

Two days later, the day of my first shift, I found myself wandering around the fourth floor to meet the person in charge of showing me the tactical ropes for the day. Her name was Vickie. Reserved, amiable, smart. She gets it. For some 10 minutes, I aimlessly shuffled through the doors. Maybe I was in the wrong place? This is where Matt's office was, right? Wasn't the IST room right next to him? 

Then my cell started to buzz — it was Vickie: "You must be the wandering fool down on the fourth floor. Go up one."

These guys have two floors in this building?!

I jogged up the two flights of stairs to find her waiting. I felt late. "Time to get started. We're on Slide sols for Oppy and we've got a Spirit sweep-and-beep meeting to get to."

Huh? Sweep and what? Oppy? … Mommy?!

————————————————

The word "tactical" is a very unique one at JPL. All you have to do is say, "I'm on tactical today," and you get carte blanc to ditch all other commitments. The MER project exemplifies this as much as any, probably as a result of nothing more than the length of its tactical timeline, which is on the order of an Earth day: "We need to get this done, and we need to get this done in the next 8 hours. Sorry [wife/husband/dog/boss/new CD you haven't listened to yet]." For the orbiters floating through deep space, tactical timelines become weeks; strategic planning goes into years. MER is never more than a few weeks ahead of itself.

It took me three long months to understand this subtlety. I eventually realized that I was (am) one of the few members of the MER tactical team that hasn't been on the mission for at least several years. It seems like most of the people have been around since the 90-sol Prime Mission in 2004. I have this image in my head of the MER demographics: If you made a histogram of the "MER age" of MER tactical team members — how long they had been working on the tactical team — you would see a huge number of people in the "since Prime Mission" range, and a smaller number of people in the "just joined" range. There's this ghost town in the middle ground. Maybe I'm picking the wrong metric, though. Maybe there's something else to it.

In any case, imagine being one of this supposed majority of the tactical team, operating on this day-to-day basis of flight operations for 2600 (coming up on 2700) sols. Gets in your head a little. Makes you want to do it every day. I've been a tactical guy — at a mere half of my working time, mind you — for only nine months, some 350 sols or so, and I can't get enough of it. Every morning you're on tactical, you want to walk in and say, "GOOD THE HELL MORNING EVERYBODY." (Copyright Scott Maxwell) You forget all about the mundane and the monotonous; today, you're playing with cars on Mars.

There's another half to the story, and that half is Spirit's. I had limited involvement with Spirit because since day one for me — July 21, 2010 — she was silent. I arrived on the team when they were just developing the infamous "Sweep and Beep" protocol. Spirit was only a legend, a myth. My relationship to her is similar to Rob's relationship to the infamous Charlie:
She's in the [expletive] phonebook! She should living on Neptune! She's an extraterrestrial, a ghost, a myth, not a person in a phonebook!
(Incidentally, when we called for Spirit, we only got her answering machine. We left a nice polite message, and we never heard from her again.)

Once I was certified for tactical on Oppy three months after I started, the team chief gave me a few Spirit "shifts." The so-called tactical timeline no longer applied to Spirit and we only needed 24-hour resource modeling. Our chief MER power engineer was eternally grateful for this work. Her kind soul always seems to think it was something we didn't want to do; I really can't understand why, given the swipe-and-paste nature of it. Piece of cake. Again with the whole me-taking-my-job-for-granted thing. 

Honestly, these Spirit "shifts" were… dispiriting. (/Paolo joke.) I had to do in the Spirit IST room — still empty, still cold, still dark. Nobody else was around. The lights were dimmed, never at full brow. All the chairs but mine were butt-less until the downlink lead would come in and build some products for the mission control team. I knew the work wasn't very meaningful. I got the feeling that this room was once buzzing with the same energy that continues one floor up for Opportunity. Now it was just sad. 

I never did a single "true" tactical shift — building and modeling uplink products for a sol — for Spirit. I never got attached to her. Oppy's my baby now and I feel as much a part of the team as ever; Spirit' is just crippled, buried, and silent. She must have been pretty cool, though, if her absence can cause the kind of disheartening looks that creep across the team's collective face upon walking into that room.

As if saying goodbye to Spirit weren't enough, the Mars Science Laboratory is in the process of taking over MER town. For 7 years, MER has had un-challenged domination some two and a half floors of Building 264. Strewn throughout those floors are the other Mars programs — Mars Odyssey 2001 and the Mars Reconnaissance Orbiter — but it's been MER town for all intents and purposes. The most stunning spectacle in all these floors was the Science Office Working Group (SOWG) Room that occupied nearly half of the 6th floor. The nickname for the room became "The Callas Palace," after project manager John Callas. He speaks the Steve Squyres language — perhaps a different dialect — and fully believes in the true integration of engineering and science teams. Take notes, kids.

The SOWG room reflects this perspective of his: a large and open middle area with a full-size model of the rovers; one row of a continuous table wrapping around three sides for the science team; and the infamous "back row" of engineers to provide reality checks and "no you can't do that because this is Mars and Mars is unforgiving to your idiot ideas" commentary; six projectors and screens lining the walls on two sides, with a touchscreen interface and connections to the off-lab teams; and pitch-black screens lining the windows behind the projectors. Special tours for the public in this room would result in gaping jaws as they watched everything work in concert. Spreadsheet display here, LTP report there, Maestro here, RP sim there.

The TAP/SIE — you know, a chump like me — controls all of this during the SOWG meeting. Once I learned how to operate the damned thing, when I pushed those touchscreen buttons and watched the whole room do my bidding, screens rolling down and computer inputs changing and phone lines buzzing, I felt like God

Oh, right, the pitch-black screens. A few months ago, the MER project got its final use of the SOWG room before handing it over to MSL. I was lucky (sic) enough to be on shift that day. I had used the room for only nine months and I still felt just the teensiest bit sad to have to say goodbye to it. After we finished up the SOWG meeting that day, we lifted the projection screens and pulled up the black shades:



I had always known they were shades because back in the late summer, I remember making the mistake of putting my hand on one of the shades on the east wall at about 1 in the afternoon and nearly burning my hand from the Sun beating down on the other side. But a sizable number of the MER tactical team had no idea that they were shades — they always thought there simply were no windows. Much of the team had spent 7 years using this room and, despite now being decidedly much less slaved to Mars time, forgot that sunlight is bad for humans when you're not on Earth time. Something obvious and clear to a newbie like me was buried down in the noise for everyone else.

I snapped the picture above and emailed it out to the tactical and science teams, subject line, "And the shades are opened…"

I got a few responses, all reflecting one of the following three:

"I had no idea they were shades."
"Makes me sad to see her go…"
"Much too bright."

I instantly — instantly, unequivocally, unambiguously — felt like an outsider to this project. I was set apart from that moment on, if not on the day I joined. I do not consider myself a true tactical MER team member for the simple fact that I didn't plug away at it for so long. Hell, I was still in high school when the rovers landed. Here were these people with this deep attachment to this room, just a room, and I felt only a hint of the remorse of letting it go. It's not that I don't feel included on the team — they have embraced me as well as I could have ever hoped for, and I hoped for quite a bit — it's that I don't have quite the same attachment. They are these strange, wild beasts. I feel like I am on the outside, looking in.

(Ahem.)

This gives me this bird's-eye point of view, allowing me to sit back in awe at this team and the way they do tactical operations. I've found it fun to intentionally ask a hot-button question that harkens back to years long gone, to a process or analysis or conversation in which people have personal and emotional stake, and then just… watch them.


What I get out of it is a useful rant, a half-furious monologue with content. It oozes fundamental rights and wrongs, fundamental strengths and flaws about how we do things. I take mental notes, as there is some professional stake for me to know these bits of knowledge.

I am now highly informed on day-to-day, non-deterministic tactical operations, sequence modeling, and — most importantly — what people mean. I am now in a position to watch, in near-real time, some piece of metal take images on — on, not around, not above — another planet. I am now one of a few dozen people in the entirety of human history who gets to say, "Yes, I play with cars on Mars. Yes, it kicks the crap out of anything else I've ever done. And yes, I don't think I'll ever get this lucky again." 

The official farewell to Spirit, now a few months back, reminds me of a lot:  the SOWG room is gone, MSL is creeping steadily upwards, and only Oppy is left to toy with. But it all seems to come down to Spirit. I never thought I'd be a rover kind of guy. Really, I feel tricked, like somebody duped me and filled my head with that anti-rover nonsense. And yet here I am. I never really touched the Spirit side of things, never got my hands dirty on the 4th floor.* But hey, you know… it's just as cool to watch this in lockstep with the rest of the battle-hardened team.

RIP, Spirit — I didn't know you, but you seemed pretty cool.

————————————————

I think it should be known what the "TAP/SIE Album of the Day" was for my last two Spirit "shifts"…


(I hope we can appreciate the naming convention there.) 

That last one, Dust Bowl, was especially apt for the moment. The obvious reference to the "dust bowl" (Mars) aside, if you stretch the lyrics enough, you get the sense that Spirit is singing this song, living in a dust bowl, her human counterparts unable to console her. Plus, it's got a gnarly solo. 

————————————————

*I have my claim to fame: My last Spirit "shift" was, incidentally, the last Spirit tactical planning shift for the entire mission. From now until the end of time, I will hold the all-time MER record for number of sols planned in a single shift: Twenty-eight. (I can feel the team saying, "This guy? Who is this guy? He got the record?" Eat your heart out!**)

**As it turns out, nobody but me seemed to care. Newton's… 11th?… Law: Matt is narcissistic.***

***Obligatory third footnote.

Wednesday, August 10, 2011

Paolo 2, Matt 0

Paolo has, so far, hit my hanging curveballs right out of the park. Twice. Two times. He can't help it if I lob it right over the plate for him:

Incident 1: 26 July 2011

Upon seeing me enter the IST room with crutches and watching everyone else give me sympathy, Paolo states the following: "So… one might say you failed a slip check?"

Paolo 1, Matt 0. Also, touché. Also, funny.

Incident 2: 10 August 2011
Upon my asking if anyone had a pocket knife to fix the loose screws in my glasses, Paolo states the following: "So… one might say the TAP/SIE has a few loose screws?"

Paolo 2, Matt 0.

Live to fight another day, Matt.
Which brings me to today's lessons:

1) Paolo is quick on his feet.
2) Paolo is smarter than Matt.

You can tell by that clever smile of his when he cracks these kinds of comments. He must know that I'm a sucker for cheap jokes, because I always laugh at his.

Which brings me to today's TAP/SIE Album of the day: 


Which brings me to today's planning summary for sols 2683 and 2854:

1) 25-meter bump to the Odyssey Crater. I smell a boulder field and images of a fresh crater. Executing... tomorrow? Afternoon? Evening? I've lost my sense of time. Need another cup of coffee.
2) Light remote sensing. We're preserving flash volume for the oodles of full-res Pancam images and IDD work that we'll we doing in the upcoming weeks.
3) My ankle still hurts.

#3 is the important one.

EDIT:

Today's featured picture, from Stu: (false color, but sweet as hell)


Tuesday, August 9, 2011

I have instituted a new project-level requirement


Matt shall grow the equivalent of a playoff beard for the Endeavour Crater campaign.
Figure 1: fig. 1 and fig. 2

My inability to grow reliable facial hair notwithstanding, it's the thought that counts. I started it two nights ago. You know, for good luck. And stuff. 

Saturday, August 6, 2011

Periscope up: Post-drive, sol 2678

Welp. 



Not too shabby. You can see the CY terrain change right on Spirit Point, and it also looks like we've "gone over the edge" -- peering right down into Endeavour.

Thursday, August 4, 2011

A Day In the Life — bSols 2676-2677 — A Change of Perspective

There are only so many places to sit around the center table in the MER-B sequencing room. A reasonable estimate puts this number at 6, not counting the ends. The end towards the door is usually John Callas' spot, when he chooses to stroll from his office just down the hallway to the room to get a glimpse of the action. The opposite end spot makes you sit with your back to the projector screens. Do not want.

Monday (planning sols 2674-2675) was a shadow-heavy day: Will, a former-and-now-returning-in-all-his-glory tactical guy, wore the TDL hat with Patrick; the infamous Brenda "Brenda Jeaux" Franklin wore the TAP/SIE hat with me. Brendajeaux — pronounced in a southern accent, in one word, don't be shy — is always quick to remind me that she and I could never get married, never mind her being old enough to be my grandmother. Otherwise, she'd be Brenda Lenda. (We'd have to put all Linda's and all Kendra's in the same category. You know, just to be sure.)

I like to stick with something more palatable: When she's shadowing me, it's Team Brenda-Lenda, in the hizz-ouse. (The hyphen is important, I promise.)

Today, though, was a wholly unseen monster: Will shadowing Patrick again as TDL; Richard (a seasoned TAP/SIE, TUL, and Navcam/Ecam PUL to boot) shadowing Mission Manager Matt; Nimisha (a two-year TAP/SIE veteran) shadowing the reserved Vickie as TUL; and a new hire, Heather, shadowing yours truly on TAP/SIE.

Oh, right, Albert was back as SOWG chair, so he plops down at the table, too. The RPs invited a summer intern to join us. John pops in and out, too. With another intern in the room, the body count is at 12. I stash my laptop and bag of food in the TAP/SIE corner, staking my spot before more tactical rabbits take it over. (And with the arrival at Cape York, we can expect more days like this.)

From my perspective, what made today interesting, and what makes recent planning days like it interesting as well, is the presence of a shadow. Shadow training is, really, the only way to learn any of the roles in tactical. There's no true "TAP/SIE for Dummies" resource. There are documents galore to get going on the basics — you know, like, here's Mars, here's a rover, here's a joystick, just kidding throw away the joystick and don't be silly, here are the dinosaur bones, just kidding we haven't found any yet, just kidding it's a conspiracy and they've managed to silence us — but those only go so far. There's one way to learn: Shove the newbie into the pool. 

There are several joke names I've come up with to refer to TAP/SIE training. My current favorite:

How to Make a Perfectly Capable Engineer Cry

It's the firehouse nature of it. It's that it is almost entirely idiosyncratic. (By the way, my desert island, all-time, top favorite underestimation of the world of engineering: Most of it is idiosyncratic. Get used to it, kids.) It's the requirement to… just… remember things. You just have to know. You just have to get it.

Osmosis is the key: take it all at face value; remember that the TAP/SIE checklist is a checklist, not a procedure. (Recent estimates of turning the TAP/SIE checklist into a true procedure place it at… one bazillion of whatever the longest book is. No more, and no less. Same for the RPs' procedure, I'm sure.) Some of it is mindless thumb twitching to power through the scripts' execution, while some of it requires actual thought. Turning on the brain in the right places is another acquired skill of the TAP/SIE.

I've found that training, on either the giving or receiving side, is like quizzing the musically illiterate. Let's pretend JJ Cale's "Cocaine" is playing, and somebody goes, "Hey! Eric Clapton!" Do you get angry? Do you blame The Man? Or do you just accept that they just don't know any better? Or, let's pretend you flat-out just ask the person, "Explain to me the significance of Derek Trucks touring with Eric Clapton." If their benighted eyes just stare at you, do you make them feel bad that they don't know? Or do you accept that they just don't know any better?

(The latter.)

———————————————

Anywho. 

Since Friday, our perspective has changed in more of a poetic sense. Since solar conjunction and Santa Maria, our Long Term Planning leads have been presenting something we call the "sol path" — a 1-week-ish, high-level plan — with the words "Drive towards Endeavour." Starting last Friday, those words were changed:

"Drive to Cape York."

This represents a subtle but significant change in our tactical stance. Most of my year as a TAP/SIE has seen only long drives across small ripples, with the occasional crater popping up. Santa Maria was my first experience with sequencing actual science, in the sense that most of the plan was geared towards the science. In the long Meridiani expanse, there is, of course, some science to be had: poke a rock here, 13-filter some outcrop there. For the most part, though, it's all been about supporting drives. This changes your mindset, your paradigm, your modus operandi

Steve Squyres sent an email out to the team to this effect. Paraphrased: "Let's take our foot off the pedal and catch our breath."

The planning for sols 2676 and 2677 took on this mindset, though there was some whiplash from slowing down. The day before, Frank, one of our RPs, gave a presentation to the sequencing team (IST) that was then a few months old. It was about the plans for the approach onto Cape York. He had mapped out two paths: Path A enters in the direction of the center of Endeavour from the west side of CY through what looks like an "opening," which is likely just a break in the surface contact; Path B starts at Spirit Point and climbs onto CY from the south. 

Both paths seem safe for the rover to drive on. However, Path A gives us a better ("less oblique," as Frank puts it) angle into Endeavour and isn't so awkward, if his estimates of the tilt at various points on CY are correct. On the other hand, Path B has some potential targets right at Spirit Point, and there's no reason to force driving back to the Path A entrance because getting to the middle of CY from Spirit Point looks easy enough. All considered, Path B was the preferred choice.

This was sitting in the back of our minds through the SOWG meeting and APAM a few hours later. Our RPs had built the basic structure of their drive and were ready to flesh out the details. That was until another RP walked in during APAM and dropped the bomb:

"Didn't Frank want imagery of Path A's entrance point?"

If we had sequenced the drive as planned at that point, we would have blown right past a decent spot to get the last shot of the Path A entrance onto CY. We were forced — though I use that term lightly —  to sequence any such imaging in the post-drive block. Both TULs and both MMs were uncomfortable with adding even mid-drive imaging to the plan at this point. Although it can be considered a trivial process, that's exactly where we make mistakes. Pre-drive imaging of the entrance was out of the question, too, since we were about 100 meters away from the entrance before the drive. 

Frank had at least wanted some images of the entrance in case we did, in the end, find that Path B was not traversable. We'd be a few steps ahead with the information from possible. We nuked the AutoNav portion of the drive — killing 10 or so meters from the drive — and opted for the usual blind drive for 120 meters. We wanted to add a second set of Pancams at the end.

But! Hey! Wait!

The trade-off wasn't only about how or when to sequence them. It's not that easy. The next variable was about preserving memory on board: we're a little bit-starved right now. We clean out the rover's flash memory pretty regularly but things can creep up on us. We've got enough room to store the additional images, we surmised, but what about downlink priority? When should we make them come down? What about Friday's planning, when we want to do some color imaging and will have 3 sols to plan? Do we want to fill flash so willy-nilly? What about visibility? Will we even be able to see it, or will the additional Pancam frames be wasted bits? 

Wait, who's decision is this? It sounds sciencey, so Albert should take charge. Well, no, not really, it's about bits and drive distance, and it sounds engineeringy, so the RPs should take charge.

A voice comes booming in from the phone speaker — it's Steve. I didn't even know he had dialed in. From the corner, I'm all like

Figure 1: what is going on

And Steve's all like

Figure 2: Kang. Kang the Conqueror. He conquers things.

He says, "How many RPs are there in the room? Make a decision!" Almost like saying, "Hey kids, put the legos down! Get the job done!"

We opt for a 4x1 Pancam at reasonable resolution. Better to have it and delete it later if we find out that it's junk.

Figure 3: For. The. Win.

For the record, APAM is usually a meeting that lasts 15 to 20 minutes. We pushed 1 hour with this one. The time didn't bother us — we had plenty of tactical margin given that our uplink wasn't until the next Earth day anyways (Restricted planning is not for the impatient) — nor should it have. There were many tradeoffs being tossed around, many of them apples-to-oranges, or apples-to-some-fruit-nobody's-ever-heard-of.

———————————————

The rest of the day went rather uneventfully. I'll take this time to introduce a new concept and new post theme:

The TAP/SIE Album of the Day

Around October of last year, when I was still in TAP/SIE training, I introduced this into my Uplink Reports. A reasonable number of people see my report, so I figured I must give them some goodies. Each new tactical shift, I list the next "album of the day." It's what I'm listening to, or what's in a playlist I'm making, or simply what's in my library. Yes, this is required listening. No, you can't complain.

The first ever TAP/SIE Album of the Day was Monte Montgomery's "Live At the Caravan of Dreams." This past Wednesday, it was Joe Satriani's "Live In Paris." I try to span my genres like a good music nut: Eva Cassidy; Erykah Badu (NSFW!!!); Black Country Communion; Paul Simon; Walter Trout; BB King; Bnois King; you name it. If it's good, honest music, I'll make it the album of the day. 

(Letters to the editor entitled, "Hey Matt here's a new album of the day for you," will be summarily ignored. I didn't say I'd be fair about it.)

It rarely grabs the attention of those in the sequencing room, even though it's displayed for part of APAM and the Master/Submaster Walkthrough. The one time it got some actual conversation was when Vickie Scarffe, resident TUL and TAP/SIE, asked me, "Is this just off the top of your head?"

"Yes," I replied, "sometimes. Sometimes not. Sometimes I've got an agenda about it. Moreover…"

"Moreover?!" said Julie Townsend, RP2 that day.

"… yes, moreover, it's what you should be listening to."