Help Solve the Mystery of the Pioneer Anomaly

deglr6328 writes "Very soon, NASA will be dismantling and scrapping its only computer left which is able to access and process the data on its ancient 7- and 9-track magnetic tapes. "Who cares", you say? Well, the Planetary Society for one and they're hoping you might care as well. The data held on these (few hundred) tapes is no ordinary forgettable data, it is the complete archive of the first 15 years of all the data returned to Earth by the Pioneer spacecraft which were sent into interstellar space. This additional and thus far unexamined data (the data after 1988 is available and has already been examined) may hold the key to solving what is considered one of the top problems in physics today, the so called Pioneer anomaly, where the observed trajectory of these spacecraft (and a couple others) deviates noticeably from our very precise expectation. The reason for the anomaly may be as mundane as uneven radiation pressure or escaping thruster fuel or it may be as groundbreaking as a clue to completely new physics, perhaps related to dark matter or dark energy. The Planetary Society is planning on recovering this data and poring over it meticulously to look for something which may have been missed or hidden from current investigations into the phenomenon. They need money to do this, about $250,000, and are asking for donations to fund the project. You do not need to be a member to donate. There are no serious proposals to send any more spin-stabilized spacecraft on solar escape trajectories any time in the near future and this is probably the only tenable method we have to directly investigate this mystery in the interim."

Re:Have you heard of Nero?

[scsirob]

There's hardly any hardware available to read these tapes anymore. Proprietary format, ancient tape drives and undocumented data formats make this a huge problem.

Pay up! It's science!

[Anonymous Coward]

I donated. How about you?

RTFF

[spellraiser]

From the FAQ [planetary.org]:

Will the data be made publicly available?
Yes. First, the data first must be recovered, validated, documented, and preliminary analyses must be done. After those tasks are completed (probably taking months to a year), the data will be made publicly available, including second-order data products when the raw data is processed by JPL orbit software.

Re:Why the deviance?

[tgrimley]

from the wikipedia [wikipedia.org] link:

"When all known forces acting on the spacecraft are taken into consideration, a very small but unexplained force remains. It causes a constant sunwards acceleration of (8.74 ± 1.33) × 10^-10 m/s2 for both spacecraft.

...

"Data from the Galileo and Ulysses spacecraft are also indicative of a similar effect, although for various reasons (such as their relative proximity to the Sun) firm conclusions cannot be drawn from these sources."

Re:Archive in different format

[mikeswi]

That's just what they're planning to do. The problem is that the current format can only be understood by a particular type of obsolete computer that NASA is about to scrap.

I had a letter somewhere that explained the problem in detail but I must have tossed it (I'm a member of the society, so I get the occasional mailing). They're planning to port the data to a modern format so it can be examined properly.

Re:Funding

[Anonymous Coward]

Because defense spending is a necessary and Constitutionally-justified federal espense. Space exploration is not. Since you really do not want any sort of national defense, we'll make sure to "throw you on the bayonettes" of the enemy next time any army, terrorist or otherwise, attacks us. Maybe your corpse will slow them down, and you can contribute in some way especially when you refused to pay your fair share. You've made your contempt for your own country and its people pretty clear to us.

What about the hardware itself?

[Colin Douglas Howell]

Does anyone have any idea what these machines are or whether they can be saved from the scrapper?

As for the data, a lot of people here seem to be really naive about how hard it is to recover old data like this. "Just download it onto a hard disk." Well, yeah, but the trick is getting working 9-track drives (relatively easy) and 7-track drives (much harder) and going through the effort required to ensure you get the data off successfully instead of destroying it. (Remember, these tapes are very old and probably extremely fragile, and you may only get one shot at recovering the data.)

Re:Have you heard of Nero?

[bazio]

The largest part of the cost does not come from recovering the data (although that will be costly), it is from the cost to actually analyze the data and perform any necessary calculations. Despite the current economic conditions, PhD's don't work cheap, at least, not the ones with the required skill sets for this analysis. Additionally, the storage cost is going to be a bit in and of itself. I am not sure of the density and capacity of the tapes NASA used, but 15 years worth of every useful piece of data (observations, navigation, status, etc...) has got to be a lot. Just "ripping it to a hard drive" would require a bit of hard drive space, and would be a bit useless, since they would just be moving it from one aging magnetic media to a slightly newer magnetic media. They are probably hoping to either move the data to some sort of optical storage, or read it directly from the tapes.

Re:Why the deviance?

[mollymoo]

Is there a Voyager anomaly?

Perhaps. We don't know because Voyager, like most other spacecraft, is 3-axis stabilised. That means it keeps pointed the right way using only its thrusters. Pioneer is spin stabilised, like a rifle bullet in flight, so requires much smaller pointing corrections using thrusters. The anomaly is a very slight one, so slight that it is lost in the uncertainty caused by the level of thruster activity on 3-axis stabilised craft.

Re:Why the deviance?

[mollymoo]

Spins like a rifle bullet, but a damn slight slower! Of the order of a few rpm.

Re:Have you heard of Nero?

[gclef]

One thing to consider: old tape is fragile. I worked with audio tape for years, and tape older than 10 years had to be literally baked (heated & cooled again) before playing. If you didn't bake an old tape the filings from the tape would slough off onto the reading heads...you might (if you were lucky) get one play out of an unbaked tape, but the audio on the tape would definitely be destroyed.

Given the age of these tapes, getting the data off without destroying it is not as simple as just slapping it onto the machine & hitting "play."

Re:What about the hardware itself?

[argent]

Sounds like a job for The Computer Museum [computerhistory.org].

I own a Kennedy 9-Track and a very old 7 track

[jbottoms]

These tapes should be readable on any midrange or mainframe. I own a Pr1me computer which should serve the purpose. Contactinformation is below: Jeffrey Bottoms, 4405 Pease #3, Houston, TX 77023

Re:How much do you want to bet...

[dlefavor]

Actually, I'd bet quite a bit.

A 2400 foot reel of 6250 bpi 9-track tape contains about 160 MB, given large block size. If they're the 3490 type tapes, each cartridge can hold as much as 1600MB. Block size is important because there's an inter-block gap on the tapes that is essentially wasted space, and the more blocks, the more waste.

This is typical of Govt funded thought processes.

[cbelt3]

I'm not surprised that NASA was unable to develop funding to convert / analyze this data. After working with them for a few years, and with the 'Military Industrial Complex" for more than that, I can clearly say:
"If it isn't in the requirements document, it's not going to get done, no matter how simple or beneficial it is."

I worked on a project back in the early 80's. We were launching missles on a test range. I was responsible for the telemetry recording. We used a massive Honeywell tape drive and a bunch of telemetry circuitry to record at 1MHz. After designing the circuitry to measure and feed the data (all analog, BTW) to the drives, I asked my boss where the specs were for the circuitry to read the data back off the tapes for analysis.

I was told there wasn't any. It wasn't a requirement. And I had better leave it at that. I kind of freaked- how the hell can we spend $100K in hardware and time to record tapes that can never, ever be read ?

The answer ? It was basically butt covering. If something happened they would ask the gummint to fund a project to read the data off the tapes.

I went ahead and designd and built a playback system on the side, nights and weekends. We went ahead and launched missles. We had guidance failures. I was asked to read the tapes. I pulled out my breadboarded setup, and read the tapes. The project team was happy, problems were solved, etc.

And I was put on the next layoff list for 'failing to obey orders'. So I got a better job, and quit before the axe fell (large defense contractor axes fell sloowly back then- lots of little clerk types had to spent their quality time with each piece of paper).

The Moral ? Never underestimate the stupidity of large organizations- governmental or otherwise.

This is a problem...

[Siener]

Danny Hillis of the Long Now Foundation [longnow.org] has been pointing out these kinds of problems for years.

Most types of digital storage is not good for conserving data in the long run. Hardware changes. File formats change. Most digital media have a very limited lifetime.

As an example: We have a very good record of the letters that Greek philosophers wrote to each other 2000 years ago. On the other had there's loads of important research data from the early days of computing that's already lost forever.

Some background...

[ear1grey]

This paper [arxiv.org] reviews the current status of the anomaly and describes how the Pioneer data could help. It may be a bit math-intensive for some, but the words surrounding the sums do pull them all into focus.

Re:It's obviously an alien plot.

[ninti]

I have been involved with the Planetary Society before and they are a group of good people. If you put a note saying this is specifically what you want your money spent on, I'm sure they would honor it.

Re:Funding

[RexRhino]

Because that is not how taxes work! If they were going to spend the money on something you really wanted or needed, they wouldn't have to forcefully extract the money from you at gunpoint, now would they? You would give it voluntarily (like people are going to do for the Planetary Society and this project).

Re:It's obviously an alien plot.

[pixelpusher220]

I have no doubt that they are a fine upstanding group with honorable intentions.

Legally speaking though, if there isn't a designation the money can be used for just about anything. I used to work for a company that did non-profit fundraising software development. It's a pretty basic requirement to allow fund designation, that's all.

I talked to the guy who used to run Pioneer...

[david.given]

...and the situation is as follows:

• The data used to be stored on magnetic tape. When the tape started deteriorating, all the data was archived off onto then state-of-the-art MO disks.
• The machine used was a MicroVAX with a DEC RWZ21 SCSI MO drive, which is apparently quite rare. The disks are 128MB each.
• For Pioneer 10, there are 155 disks, making 19840 MB of data.
• For Pioneer 11, there are 217 disks, making 27776 MB of data.
• Each disk takes about 10 minutes to read to the MicroVAX, and then more time to move across onto a real computer, of course.

I would have happily volunteered to spend a couple of days swapping disks in order to salvage all this lot, but alas, I'm the wrong side of the Atlantic. The guy in charge has recently been made redundant, and he was desperate to find someone to hand off all this to... but there's incredible beaurocracy. (I gather all the data was actually supposed to have been destroyed some years ago, but through some 'oversight' hadn't been.)

Alas, I don't have permission to publish his address, but I'll put him in touch with the Planetary Society on the off chance he doesn't know about this.

Interestingly, for years he ran the Pioneer spacecraft off a Mac Quadra 950! Check out the screen shots [nasa.gov]...

Re:But how huge?

['nother poster]

Yes. 16 bps on an 8W transmitter. At least for the pioneer 6-10 series if I read the data correctly.

Re:What about the hardware itself?

[Colin Douglas Howell]

It seems at least a couple of important people there know about the situation, judging from this recent message [classiccmp.org] on the cctalk classic-computing mailing list. (Sellam Ismail is the museum's software curator. I don't know if Al Kossow actually works for the museum, but he's certainly contributed a great deal [trailing-edge.com] to preserving computer history.)

Re:RTFF

[DrJimbo]

Rich0 said:

Yes, but the only urgent part of this project is recovering the data.

In theory for far less you could simply recover the data, test that it was recovered properly, and then stick it on a webpage for anybody in the world to analyze.

TFF said:

Will the data be made publicly available? Yes. First, the data first must be recovered, validated, documented, and preliminary analyses must be done. After those tasks are completed (probably taking months to a year), the data will be made publicly available, including second-order data products when the raw data is processed by JPL orbit software.

I am a scientist and I have worked on analyzing large data sets such as this before. The tricky part is what you describe as "test that it was recovered properly". This is what TFF described as "preliminary analyses".

It would be foolish for a programmer to publish a program on the web without first running it a few times to catch bugs. In the same way, a scientist must check her data (even "raw" data) before just blindly putting it up on the web for all to see. If she posts faulty data then she wastes everyone's time, she looks like a fool and pisses a lot of people off. Her reputation may be ruined.

But how can she know for sure that the data was recovered properly? Checking parity bits is not nearly enough, because she needs to know for sure that she did not make any subtle mistakes and that no one in the chain of generating and producing the data made any subtle mistakes.

One necessary (but not sufficient) step is to actually analyze the data with your model(s) and see if it makes sense. If it doesn't, then you may need correct your transcription procedure and go back to the original tapes and read them again.

Transferring the data for $10K and not doing the preliminary analysis would be foolish beyond belief. I think a better cost estimate is roughly $100K for the transfer(s) and preliminary analysis needed to ensure the transfer was done properly. Since everything would already be set up to analyze the data it makes perfect sense to also get another $100K to do the "real" analysis. Since something may well go wrong, ask for another $50K so you can be reasonably sure of getting it done right in the first go round.

pioneer data all available online

[intelsquirrel]

Here are some links published on another group about this same topic, and all data that NASA knows about is already saved.

> http://nssdc.gsfc.nasa.gov/database/MasterCatalog? sc=1972-012A&ds=* [nasa.gov]
> http://nssdc.gsfc.nasa.gov/database/MasterCatalog? sc=1973-019A&ds=* [nasa.gov]

>Well, well, well...it looks like every bit of Pioneer 10 and 11 has been saved already, and can be accessed thru the proper channels (on tape, but apparently they will burn a CDROM on request).

Re:Lots of other data

[Ignignot]

And based on my extensive personal experience, anyone who expects a couple decades worth of knowledge that WE ALREADY HAVE in other formats (ground based stations, etc) to prove or disprove something as complicated as global climate change (and very sudden, non-linear global climate change, to boot) is going to be disappointed.

If you are interested in a history going back to at least the 50's, with extremely good resolution, just ask the NOAA for it and they will be happy to furnish you with more information than you will know what to do with.

I know the parent or grandparent are probably trolls, but this deserves a useful response.

Re:Why not?

[DerekLyons]

Why is there a big hoopla about Planetary society raising meager $250,000? You need money to carry out research. NASA obviously doesn't care much about basic sciences, and is quite busy wasting tax dollars in 'spectacular' but dumb and useless shuttle launches.

NASA cares plenty about basic science - and the basic science these tapes were meant to accomplish was accomplished decades ago. (In fact a great deal of both the data and the science is available on the web.) What TPS wants to do is 'extended science', something NASA has been trying to get funding for over the past few years - but has been unable to do so.

TPS is essentially lying to you when it claims the tapes haven't been read or analyzed.

Those old tapes may already be unreadable

[DrTime]

I worked for the Lunar & Planetary Lab many years ago when we recieved hundreds of such tapes from Pioneer. At the time they were stored in racks in the computer room with an IBM 1130 system. Track density in those days was either 556 bpi or 800 bpi on 1 inch tape (memory) and recording technology was crude at best. Even then the fear was that the data would be irrecoverable in years. At the time, the lab was run by Dr. Sonnett who credited with the discovery that CMOS circuits were static sensitive. He came up with idea of grounding workers. This paved the way for low power electonics on some of these payloads. They were interesting days.