Sunday, January 29, 2012

100 Sols on Cape York — An Absurdly Brief (Read: Long) Summary (Part 8/8)

Part 8: Where we stand — and where we're going tomorrow

This, the belated Part 8 of my "100 Sols on Cape York" series, is late for a few reasons:

1) I fly lots of spaceships and they all started getting jealous. I keep telling them, "No no no, see, MER and I just friends."
2) I was waiting for Opportunity to turn 8 years* old so it could look like I was being cute — "8 years for Part 8… Get it?! Get it?!" **
3) There hasn't been a whole lot going on.

There actually has been a lot going on, but it's all invisible to the folks outside of tactical. Opportunity won't be moving for a good long while; we've got nearly 15 degrees of tilt in the northerly direction, plenty to allow us to survive the depths of the Martian winter while still collecting good, quality science. Opportunity's current view is summed up nicely in what we're calling the "Greeley Pan", named after the current location (Greeley Haven) and the instrument collecting the images (Pancam).


(Click for full-res awesomeness.)

Our team estimates that once we get past winter solstice on March 31***, it'll take a few months to get back to where we are right now in terms of available power. 

So, Matt, until then, what can Opportunity do in the meantime? 

We can certainly do everything that Opportunity has up her sleeve: driving; IDD'ing; remote sensing; backflips. We just can't do a lot of it. 

What are your real science collection options, and what are their relative priorities?

Astute question! The activities are, in decreasing priority, as such: 

1) Radio Science DTE ("RS-DTEs"): By analyzing the signal coming from Opportunity's High Gain Antenna (HGA), somebody way smarter than me can figure out how "wobbly" Mars is. It's much more complicated than the word "wobbly," but I'll settle for it. The wobbliness of the planet will cause the Doppler shift of the HGA signal to vary in predictable ways; a totally solid planet would wobble in a very different way than a planet that isn't so solid on the inside. The math to describe these variances is well known so all we need to close the error bars on what we already know about Mars' interior is a stationary radio source on the planet, or at least a radio source whose position is very well known. Hence, Opportunity is a prime candidate for this kind of measurement. Brilliant!
2) APXS+Mössbauer investigations of Amboy; MI Mosaic: Right at our feet is a bit of outcrop. On it is a nice smooth surface that is itching to be poked at by the IDD. One of these spots on this surface has been named "Amboy." The content of Amboy will help the scientists know what's going on at Cape York. Combinations of APXS and Mössbauer measurements will tell them all sorts of cool things about it. And, hey, since we're there, why not use the Microscopic Imager (MI) to make a nice mosaic? Exactly! We're currently queuing up a 20x2x5 (20 images long, 2 images across, 5 images deep) mosaic. It'll probably be more like 17 images long, depending on reachability.
3) Greeley Panorama: See image above. This is being taken piecemeal because of the sheet amount of time and data it uses, with similar lighting conditions across the board. A funny "feature" the Pancam team will have to work around is that we have sequenced most of the Greeley Pan segments in the morning science block (or "AM block"), when the HGA may be articulating as it tracks the Earth for the uplink communications passes every morning. You can see in the image above that we did indeed capture, in two Pancam images, the HGA in two different places within several minutes of each other.
4) Everything else: Your usual collection of random remote sensing observations. We have weekly/monthly/whatever-ly liens that we still like to cover.

Yours truly is one degree removed from NDT as of last night. This is what I like to call "cool."

So we can move, but we're not. Really, why aren't we boogying?

At our feet is a potentially rich source of science targets and we would like to be able to skirt along it and IDD it all to death, but we can't. Errrr, shouldn't. Why not?The primary reason is, of course, available power. This is the over-arching constraint that drives what we do every day. However, there are two benefits we gain from not moving.

1) RS-DTEs
2) AXPS+MB investigations

As I mentioned above, RS-DTEs require that we be in the same place; if we have moved, we have to know exactly where and how much down to, roughly, the centimeter level. These things matter when you're measuring single-digit Hertz doppler shifts. Using that nice little piece of flight software called VisOdom, we can track rover movements to that accuracy. However, that doesn't mean we should move. We want to completely characterize Amboy with the IDD and any movements would mean Amboy might not be in reach. (This is certainly true for azimuthal/rotational movements, and could be true for translational movements along the surface. Think about Oppy's damaged IDD shoulder joint and you'll understand why.)

Therefore we have incentive to stay put. The only thing a small "bump" drive would gain for us is some extra northerly tilt. We can get another couple degrees if we want it because the nearby terrain has it for us. However, at 15 degrees, we're doing alright. Better is the enemy of good.

It sounds like things aren't so busy if you're scheduling only extra communication windows and Pancam sequences.

Couldn't be further from the truth! 

We're living sol to sol on the power analyses. We're not really cutting it very close, but we love to be conservative. Think about our constraints over time: Every bit of the campaign on Cape York has been constrained for one reason or another. Previously, it was downlink volume and knowledge of the state of the rover, or it was the successful collection of APXS data, or it was [insert thing]. Now, we're living sol to sol on power constraints. Every day, the TAP/SIE (read: me, sometimes) gets to take the activity plan for the next sol(s) in the plan and model it. Using lots of big fancy math and computer simulations and things that look nothing like Hollywood would portray it, we come back to the team and tell them what we found. 

In technical terms, this verbal report ranges from, "Yah, we're cool," to, "We're cool if we pull out the [observation]," to, "Nah, we're going to have to do some surgery to this plan," to simply, "Nah." Every day, the team lines up the priorities and we allow the science team to pack it on. From there, we trim — it's easier to take away than to add, as it turns out. 

There are many permutations of combined activities that we can bless with the canonical "this won't kill the rover" stamp. We also have to be mindful of strategic (long term) decisions while we do this tactical decision making. This permutation process takes several hours, so even though there really isn't that much science activity going on, the days are still mind-bending and long.

Lately, this permutation process has resulted in a few interesting patterns of activity on the rover. For instance, for a given sol we have to choose between doing a UHF downlink relay pass with an orbiter and doing an RS-DTE. This is apparent to the amateur MER follower when there are a few days between image data on the ground when they expect to be getting some data every day. Because the data we're collecting is low-priority for rover health and tactical decision making, we can afford to not downlink the large (I speak relatively) amounts of data from the UHF relay. RS-DTEs give us a tiny snapshot of rover health, and although they don't tell us much, they tell us something useful. We can usually move on planning the next sol(s) with just the DTE "snapshot."

Another pattern is the placement of the Greeley Pan observations. The more perceptive MER data followers will note that most Pancam Greeley Pan images come down the sol after they were collected. This often happens anyways when we're tight on data and the low-priority Pancam images get shoved to the bottom of the downlink queue and don't make it down for a few sols. However we're not in this data-tight situation. Why, then, are they "late"? A sol's sequence of events will tell us why:

[ Start sol's plan ] >>> [ Do stuff ] >>> [ UHF Relay ] >>> [ Nap overnight ] >>> [ AM Block stuff ] >>> [ Next sol's plan starts ]

We straddle the local midnight barrier for each sol's plan. To save on power on a given sol for the UHF relays, RS-DTEs, or APXS/MB observations, we've removed most of our remote sensing blocks from the early afternoon before the UHF relay. Instead, we make more use of the AM block. This gives another advantage in that the time of day of the AM block is roughly the same, so shadowing and light in the images will be roughly the same for the whole panorama. The AM block of a given sol's plan occurs on the next sol, and this is some time after the UHF relay has occurred. In fact, anything that is after a sol's UHF window won't get down to the ground until at least the next sol,. maybe later. Therefore, many of the Greeley Pan observations show up "late".

———————————————

This sorta, maybe, possibly wraps up the 8-part series. Right. We now return to your regularly scheduled MER blog.

—————————————————————

*Surface years. They're like dog years, with the exception that they're the exact same length as normal years. So, normal years… on the surface.

**Mike the Mission Manager's idea. Though nobody reads my footnotes so he'll never get credit!

***Solstice is actually a few weeks after minimum solar insolation (insolation = amount of photons hitting Mars) at Opportunity's latitude. Insolation at the surface is defined mostly by two parameters: distance from the sun and tilt relative to the orbital plan about the sun. Solstice is defined by the latter (tilt), which doesn't have to coincide with the minimum insolation, though it does very closely. If you're of the number-crunching type, crunch those numbers and tell me if you agree! I know you will; as Ben Goldacre says, "If, by the end, you reckon you still might disagree with me, then I offer you this: you'll still be wrong, but you'll be wrong with a lot more panache and flair than you could possibly mange right now."

Tuesday, January 24, 2012

"8 more years, starting tomorrow."

Tonight at about 9pm Pacific (roughly 13:15 Local Solar Time at Meridiani), Opportunity will have spent 8 years on the surface.


Our SOWG chair put it best: 


After having spent 8 years doing surface operations, we need a recharge sol.

This is, of course, in reference to our current ambitious science campaign and trying to do it with a very dusty rover in a very cold winter. More on this in Part 8 of the previously-promised-I'd-have-this-done series of our time on Cape York.




(djellison @ UMSF)

Friday, January 13, 2012

100 Sols on Cape York — An Absurdly Brief (Read: Long) Summary (Part 7/8)

Part 7: Tryptophan Tilts at Turkey Haven

Sols
2772 to Present
"tl;dr" Description
We left Homestake for a haven or two.
Highlights from Matt's Notebook
Sol 2778-2779:
> The DSN is helping look for Phobos-Grunt. Might steal our uplink. &^$*#&$#^%
Sols 2783-2784:
> RPs wanted to do something fancy. TUL said "no." RPs said, "Ok."
Sols 2792-2794:
> Secret Woids: boggle / inbred / accost. I want better ones.
Sol 2800:
> AGU Conference today. Broke the news on Homestake. Can finally finish that blog entry.

With Homestake finally out of the way, we had about a week before the Mars Science Lab launched and had great potential to steal our time with the DSN for a good while. (A thing that, incidentally, never occurred. We couldn't have asked for a better launch from MSL.) Without time on the DSN, we can't uplink; without uplink, the rover can't move; without movement, she'd be stuck in a region with poor northerly tilts. We needed more than 10 degrees of northerly tilt to be safe and still perform science observations over the winter; 15 would be better.

We sequenced a nice set of drives (with a dogleg around the icky — although did that icky ever turn out to be icky?) to get ourselves to a standoff/scout position where we could image two spots where we knew there were good northerly tilts according to orbital imagery and subsequent digital elevation maps. We needed the ground imagery from this standoff position to resolve the orbital measurements and provide some much-needed precision.

(Sol 2777)

We found Turkey Haven to our south, with tilts a little above 10 degrees in the northern direction and bits of outcrop, to boot. Here's an image looking south on the last leg of the drive indicated in the image above:



From both Turkey Haven and the previous standoff position, we could see another bit of northerly tilts to the north: Winter Haven. We had two places to sit, each of which had good tilts and perhaps some good science targets. We also knew that these were the best two sites for the tilts because the ground-based imagery validated the orbital imagery and gave us confidence that the sites the orbital data showed were indeed the sites we wanted. No need to turn 'round the northern corner of Cape York to see if there was anything else that the orbital images missed. To get an idea of the tilts, here's a great virtual overlay of Oppy from Stu:



With the image-mapping of our winter sites complete before we felt like they had started, we had some time to ourselves: we selected a smooth part of the outcrop under our feet at Turkey Haven to poke with the IDD. We called it Transvaal — and here's an approach+Hazcam shot of them:

(Sol 2785)

"Hey, whaddya know, the power went up a few percent with that tilt," we told ourselves when we bumped up onto Turkey Haven. Hey, you know, that's how reality works.

We poked and Transvaal and the surrounding outcrop for roughly a week. The results? Nothing juicy. Same rock, different sol. Gave us no pieces for the Martian water puzzle. With more targets at the other Winter Haven site, a few meters to our north, and with better northerly tilts there compared to Turkey Haven, there was no reason to stay and give Transvaal what it had coming. Time to move on to Winter Haven. 

Somewhere along this whole trip, time... disappeared. Just went away. Magically, we were at another 100-sol boundary. In the sequencing room at JPL, we post a sign at the door that tells everyone for which sol(s) we're building commands today. I may or may not have stolen one of them:

Not stolen.

I dibbs'd it (a dibbs is valid even if nobody hears you!) and took it. Felt like it was cool to have one of these. Felt like I'd be keeping it around, even if 2800 is just… a number. I wonder who'll get sol 3000? (5000? 10000?)




——————————————

This is, really, the end of the first 100 sols at Cape York at the edge of Endeavour Crater. As of the time of writing, we've tacked on another 50 like it ain't no thang. The final part of the promised 8-part series will be a summary of what has happened since then — and where we plan on going once the winter is over.


Wednesday, January 11, 2012

100 Sols on Cape York — An Absurdly Brief (Read: Long) Summary (Part 6/8)

Part 6: Drawing Blood — Homestake

Sols
2763 to 2772
"tl;dr" Description
We found Martian drywall.
Highlights from Matt's Notebook
Sol 2763:
> We found two veins, one that is easier to reach and one that is harder to reach. The latter is a "richer" target. 
> Science says that these veins could be one of the more important targets at Cape York, if not the mission. 
> I don't get the excitement over the veins. Then again, I'm not a geologist.
Sol 2771:
> Homestake ate our winter haven mapping margin. MSL launch looming. Caffeine supplies dwindling.
Sol 2772:
> These are good times.
Sol 2773:
> The vein is Gypsum. Science is ecstatic. I want to blog about this, but I can't yet. Feels like a secret club.
Sol 2775*:
> Saturday planning! John bought pizza!

Details ("the deets")
When we first stumbled on the vein (which we named Homestake), there was this strange disconnect between the reality that I had been seeing and the reality that the science team was pushing. We had bought all this extra time hauling booty across Cape York's spine in order to go to the northerly slopes — which were now on the order of tens of meters away — to start mapping them with local imagery, and yet at the same time there was this very strong push-back from the science team. There was this vein, you see, but I'm not a geologist and I didn't know what it meant. Even before we knew what it was, the science team had some idea in their collective head what it could be and it would mean — whatever it was.

In all honesty, it was this disconnect from reality that had me asking myself, "Why would we stick around?" I didn't see what they saw.

But, alas — more push from them to stick around. We, the engineers, yielded what amounted to all of our planning margin time to stick around. We needed APXS and MI, more APXS off to the side, some Pancam 13-filter images; we needed everything up our sleeves in a matter of a few days, and we needed to fit it within a strategic plan. (Which, by the way, is hard. Most of our time spent in meetings that we organize to discuss strategic issues are spent trying to figure out what the question is in the first place; the rest is cake after that. You have to wade through the abstract and the absurd and the speculative for a good hour an a half before someone and speaks up and says, "Oh, so the 1-sentence summary is that we want to know… [thing]." And then everyone is on board.)

We let the science team have their fun: We gave them their APXS and their MIs and their Pancams. From this, a fantastic view of the vein:



And they came back with these fantastic results: Gypsum (drywall, sort of). Meant something awesome, apparently. Our media relations guy came into the room — "Scott, don't tweet about this yet; Matt, don't blog about this yet. There'll be a press release, but not for a while." In the end, we could tell ourselves that everything was fine, that the detour was worth it, that the margin was all for worst-case scenarios anyways.



The only reason this works on MER is because of the size of its tactical timeline (days). For landers, orbiters, and everything in between, it's usually only engineering and flight system health and safety that operate on a responsive timeline. Why? Because decisions that must be made today are usually to save the spacecraft from harm, which means that the science payload takes a backseat — even though the flight system was built to fly the payload in the first place. First order of business is having a flight system that can support the payload system. However, with MER, scientific results can be processed very quickly and fed directly into the tactical timeline. We are "closed-loop" for both the flight system and the payload, something that is purely the result of pressing the reset button every day and saying, "Ok, so, now what?"

Which is pretty neat.

Another benefit of this tactical operations concept is that the engineers are directly hooked into the science. We attend their meetings and they attend ours. I don't understand their 7-dimensional plots and their phase diagrams and their crater impact analyses, but I do understand their meta-analyses (like, "This means watery bits were here," or, "This means that this was a really giant rock, like really really really big"). It's cool when someone you've never met dials in to the SOWG meetings in the morning and lays out their 1-paragraph justification for the observation they just requested. They're usually to the point and on the money — with some wiggle room when we translate from "scientist talk" to "engineer talk."

With the western detour to Homestake on the apron of Cape York now done, we could start mapping the purported north-facing sites on the north end of Cape York: our Havens.



————————————————

*I included a sol outside of the arbitrary timeframe of the Homestake investigation because Homestake caused this one. When we ate our margin away, we instituted a Saturday planning day to make up for a lost sol or two.

Monday, January 9, 2012

100 Sols on Cape York — An Absurdly Brief (Read: Long) Summary (Part 5/8)

Part 5: The Trek North — Shoemaker Ridge

Sols
2735 to 2763

"tl;dr" Description
Power started to look tight. It was time to go find those north-facing slopes.

Highlights from Matt's Notebook
Sols 2756-275:
> Science wants a "quick" detour to the western apron of Cape York. This will cost us 2 or 3 drive sols, at the most.
Sols 2760-2762:
> LTP reports that we haven't found much in the way of science on this detour to the northwest. Tosol we will hit the apron dead-on and turn NE.

Details ("the deets")
Here, traverses became important and meaningful again. We'd been hanging out at Tisdale and Chester Lake so long we were, for all intents and purposes*, as the same spot. As such, this part of the story is all about traverse maps and miles under our feet. The hunt for tilty goodness led us in the direction of Shoemaker Ridge, a newly-named formation that forms the backbone of Cape York.



We were in such a hurry that we actually had to forego some systematic and detailed observations of the Ridout boulder perched on the eastern edge of the small Odyssey Crater.

I'll summarize from my own words:

The primary reason for the trek north was, of course, to find those slopes that give us good sun angles for good power. All estimates point to Opportunity surviving the winter with plenty of margin. However, margin can easily be eaten away by the unknown unknowns. With that in mind, we tried to reason how quickly we'd need to find a parking spot. January seemed a good last-minute date to aim for, if push came to shove. With that date in mind, we considered one other major factor: The launch of the Mars Science Laboratory on November 25.
The view periods of the MSL spacecraft after launch mean that they happen to just interfere with all other Mars missions. Immediate post-launch priority for DSN coverage of a spacecraft is very high; Opportunity, being much older and of a much lower status for requesting DSN time, gets the short end of the stick along with everyone else. This means that we may very well lose the time we've already scheduled with the DSN. Some more strategic analysis showed that if we did not finish mapping Cape York's northerly slopes by the time MSL launched, December and January would not provide us enough time to do that mapping because the incoming solar power would be too low.
We had made it clear: Thanksgiving or bust.

There is some confusion about why we went to the north for power. The reason is because that's where the north-facing facing slopes of Cape York happened to be, and it had nothing to do with them being closer to the equator where there is more incident solar energy on our arrays. In fact, as far as some of our analysis tools are concerned, we're in the same place that we landed in 2004**. Even over some 30+ kilometers, there is no discernible difference in solar power than we can measure, all else (like tilt) being equal — and we can certainly expect the same for the few hundred meters that Cape York spans north to south. We also looked to the south towards Cape Tribulation, where there were even better north-facing slopes; they were too far away for us to make it in time for winter.

It's worth keeping in mind that our long-term strategy is go south towards Cape Tribulation because the strongest signals of sciencey things (smectites) were seen from orbit near its peak. The MER team — both science and engineering — kept this concept at the front of the conversations concerning where to park for the winter. After some long and fruitful discussions, it was obvious that Cape York hadn't given up all of its fruits yet. It was an entirely new type of formation with clues of some real juicy stuff at the Tisdales and Chester Lake. Putting on our 20-20 hindsight glasses, we now know that to go south for the winter would have meant a big missing piece of the puzzle that tells us about water on Mars. Without those glasses, we had to make a decision. We went north to the small "lily pads" that the orbital imagery and elevation maps identified as having 15 degrees or more of northerly tilt. Some of these lily pads were at least the width of the rover and there appeared to be some other targets worth our time, such as a nice set of outcrop or a perch spot for a high-resolution Endeavour Crater panorama. 

Backing out for context: Our arrival point at Cape York, by now some 60 sols ago; the suspected smectite goodies from orbital observations; and the suspected location of the north-facing lily pads for winter hunkering.



We greased the wheels up (literally, actually — we heat those suckers to get the lubrication flowing) and boogied on out of Chester Lake, climbing a very slight hill and crossing over the region of negative northerly slopes to the region of positive northerly slopes. The next half dozen or so drives along the Shoemaker spine were relatively uneventful, with the exception of a jut to the west and back to capture a small indent:



This area in the green is, in a way, a dead zone of Cape York. Not a lot of interesting things going here. At about this time, the science team started looking closely at those orbital images and making hypotheses: maybe there was an exposed "vein" in the rough? maybe we could snap a few images as we see them? maybe we could even poke at them? The RPs had made such good time cutting along Shoemaker Ridge (that green area) that we had bought ourselves a week (in sols) of margin for mapping the north-facing slopes with local imaging by the rover. 

Whaddya know, we found a good vein as we approached the western apron on Cape York (here, I maintain the same labeling for context):



As seen by Oppy on the approach:


Mmm. More in Part 6, a part I never really did in detail the first time around.

————————————————


*Not, apparently, "for all intensive purposes," as I found out probably two years ago. It unnerves me that I didn't know this for the first 23 years of my life. I can't tell if this is on the same level as an Ecam PUL not knowing who Robert Plant is, or a Mission Manager telling me they thought "God bless you" was actually something like "kuh-bless-ya" until they were 12.

** Blasted typos. "I assume that is because we landed through some weird subspace wormhole and went back in time right?" Mike the Mission Manager is ever watchful.

Saturday, January 7, 2012

100 Sols on Cape York — An Absurdly Brief (Read: Long) Summary (Part 4/8)

Part 4: "… and then we have a RAT hole" — Chester Lake Gets It

*********************
Hey! I'm back! Spent a few weeks back in Colorado for the holidays and am now back in the groove. Time to continue where I left off: Chester Lake. At the time of writing, we're almost 100 sols beyond this part of the story — and 150 since we hit Cape York. 

Must… finish… this… story… 
*********************

Sols
2709 to 2735
"tl;dr" Description
We pounced on a flat outcrop and hung out — for a while. We even Mossbauer'd!
Highlights from Matt's Notebook
Sols 2719-2721: 
> Today's Secret Woids: Fran / animosity / Bobbie. I fear for their relationship because of the animosity between them. (*rimshot*)
Sols 2722-2723
> Squyres: "We've been preserving the RAT for this. It would be foolish to leave."
Sol 2730:
> How many RPs does it take to change a light bulb?

Details ("the deets")
I didn't have a lot of tactical shifts between leaving Tisdale-2 and our next target. For about a week, I had this notion that the rover was faced one direction and the Chester Lake was at some spot on Cape York, only to find out that I was about 90 degrees and 20 meters, respectively, off. I blamed my busted ankle, which, in retrospect, seems a pretty fair judgement (as far as unfair judgements go).

After giving Tisdale-2 a taste, we started hunting down something smooth and representative of Cape York. We were here to get a taste of Cape York so that we could learn a little about Endeavour. We found just such a target: Chester Lake.


In particular, we focused on a nice smooth spot that we called Salisbury-1 (Salisbury-2 wasn't as appetizing, apparently), circled in pink here.

Now, context: a perspective shot on Cape York, vertically exaggerated by 5 times, looking northwest-ish (thank my Google-image-search-"RayArvidson"-until-my-fingers-bleed talents).



I have layered in my own annotations here, most notably the direction that the front of the rover faced for the Chester Lake campaign, to give an idea of where how Oppy was oriented. To the northwest: a feature called Shoemaker Ridge, which forms the spine of Cape York running roughly north-south. To the northeast, the inboard side of Cape York and the effective edge of Endeavour Crater. Lots of interesting things all around us.

And just to convince everyone that we were there, the MRO spacecraft used it HiRise camera to find Oppy perched right where we thought we were — we call it the "ultimate locator"!

Hey!

While we started thinking about taking a poke at Salisbury, we queued up a high-resolution Pancam mosaic of Endeavour — hey, whaddya know, our first Endeavour Pan!

Note the pretty. (Click for full-res.)

Suddenly, we were right on top of Chester Lake and wanted to get going. Thus began an intense IDD campaign that lasted longer than we expected. We started with a basic MI+AXPS attack on both Salisbury-1 and −2, settling on a targeted campaign of Salisbury-1:


After the APXS results came in, there was an itch to leave and head towards Shoemaker Ridge or more due north. As my notebook indicates, Squyres (and, well, everyone else), agreed that we should dig deeper — literally. We queued up the RAT and got her grinding. After a seekscan (where it finds the surface), a brush, an MI+APXS set, another brush, and another MI+APXS sequence, we had a nice clean RAT hole:


The Mossbauer spectrometer on Oppy's arm is nothing but a dim light bulb — it's radioactive source has a half life that was selected based on a 90-sol mission, wouldn't you know — but we gave it a try for a few sols, setting it down in the RAT hole.

Pretty pictures and the slew of MI's aside, the Chester Lake campaign was a pivotal point for the MER team. Chester Lake had long-term effects on our rover and our strategic planning that we never planned on, causing some unpleasant hiccups in operations. First, there were the anomalies beyond our control. As with most things for MER, the concepts are simple but they tend to interact in a complex way. Remember that our tactical timeline is on the order of days; what we do today affects tomorrow, directly and drastically. Then, remember that we rely on an entire system of elements and pieces to maintain this tactical timeline. When one element fails, the tactical timeline is altered beyond our control and we have to adjust very quickly. It's not a metaphorical house of cards or a game of Jenga, because that would mean that we literally couldn't operate Opportunity if an element failed. Obviously, this is not the case. It's just a very fragile system. When it works, oh boy does it ever work. When it doesn't, there's a bout of sweating and legwork to get back to normal.

While at Chester Lake, one of the orbiters circling Mars couldn't relay the data that Opportunity had collected because of its own problems. We the MIs, RAT data collection, and MB integrations going on, we needed the data back each and every planning day in order to make good time and get out of there. When this couldn't come, and when we started losing uplinks and downlinks with the Deep Space Network itself due to the orbiter anomaly, things slowed to a halt: we adjusted and added hours on the MB; we used runout plans; we turned single-sol planning days into multi-sol planning days; we uplinked early and often. Sols 2720-2730 were a blur to me: lots of shifts, each of which was unique and different because of the new anomalies presented to me as a TAP/SIE. All told, we "lost" about a week's worth of sols because of the delays in data acquisition and downlink. I use the term "lost" loosely, because it didn't prevent us from collecting more science or stop us in our tracks. 

The second hiccup hit us very suddenly: Winter was approaching. Solar insolation — and its proxy, the wattage generated by Oppy's solar arrays — wasn't looking terrible, but it certainly wasn't looking good. She's a dusty rover and she hadn't had a cleaning in a few months. The engineering team threw together a detailed analysis and projection of the rover's power state for the winter that had an obvious takeaway: find any slopes that faced north and get there. Quickly.

So we boogied to find the Promised Land of Northerly Slopes. Which brings us to Part 5...