"had the impression (uh, no quotes, sorry) that one of the primary problems was not simply pumping blood, but pumping the right amount of blood.
>A natural heart adjusts depending on the needs of the body - for an artificial heart, is this a big issue? Does anyone know how it is addressed, in general, and in this specific case?"
One of several approaches to this is to use a pneumatic "volume compensation device to allow some variability in the area behind the pump for filling. The body has a series of checks and balances, but for flow regulation it's mostly based on venous return: The arteries can constrict causing the blood pressure to rise in regional or overall (systemic) fashion, but there is precious little that will cause the veins to constrict. While unanticipated venous dilation would cause problems, generally, if the physiology is, overall, intact, this will _not_ happen.
So, what I'm suggesting here is that the flow could well be regulated in several ways. I will have to do some research on the Abiomed device in its current state, but the pump could well have a central aortic pressure sensor and attempt to pump more blood (ie., faster) to maintain a mean pressure; the pump could be triggered to "fire" when a certain fill point in the ventricular chambers is reached, or the pump could be set to a fixed rate and simply pump the available blood at the time it's supposed to pump, little or lots, depending on venous return.
Having been out of the game for a while I'm not sure anymore which of the methods is prevalent, but the first 2, in isolation or combination are physiologically attractive.
I spent a number of years in the articifial heart field while at Texas Heart Institute. It really *IS* big news.
All the artificial heart implants in humans have been tethered implants. The patient has had to be tied to a console or power source, electrical or air, with no more than brief respites of untethered activity. The devices have been harsh on quality of life, and a whole host of physiologic functions. They have not allowed normal interaction with other humans, and those few patients who've ventured outside the walls of the hospital were making brief visits away, not returning to the world.
The Abiomed pump is small enough to reasonably be implanted, and reliable enough to expect it to work well. The centers selected for the initial implants have sufficient experience with animal implantation, AND various human procedures of a more mundane variety, to expect them to be able to manage the patients well, indeed.
We're about 9 years behind where I thought we'd get to with a really viable, implantable heart, mainly because of the costs necessary to support this sort of research. It's long overdue.
I suspect that the 125,000 potential patients Abiomed cited in the article may be an understatement. Doesn't matter. If this allows some patients who were dying waiting for a transplant...or who were deemed not good candidates for the scarce resource of a donated heart... a shot at a good quality life and a time extension, this is WONDERFUL news.
My step-daughter's now 16, and a bit more "worldly" than I'd like to admit. But she's had net access for the last 5 years, pretty much freely, and virtually unfettered access for the last 3.5 years.
I do networking for a living. I've periodically dropped a sniffer on her stream and captured content to see what's going on. My wife and I have discreetly used the garnered info, when we've felt we had to, to forestall some potentially hairy problems.
I *don't* want to admit that she's going to have sex as a teenager. But, you know, overall she's a pretty sharp kid, and we do NOT monitor her activity on the Web tightly. We have not had the need to do so for some time: her judgement and her decisions to come to me with 'Net-related questions and problems have indicated she is pretty safe on the Web all by herself.
We had the "It's a mean Net out there" talk back when she was first starting, and again about 18 months ago. She knows she can always come to me with stuff regarding the Net and the Web. And she has, even coming for advice when a friend of hers started having someone spoof her mail to friends in a particularly offensive way... and we nailed him, too.
I tend to agree: Trust and time are the keys. And while I've monitored, I've also trusted... Consider the monitoring a form of verification, and realize that we take our role in shaping her life pretty seriously.
Actually, the NTIA assessment of this stuff is that, Yes, it may work, and, based on actual and theoretical evaluations, its potential for harmful interference is significant above certain threshholds for power and "chipping rate."
The ultrawideband nature of this approach appears to offer a splendid opportunity to raise the noise floor across a large region of RF spectrum, to the point that little things, like GPS navigation and cell phones as we know them today will be adversely affected.
You're right: It's a neat development. But has it been appropriately evaluated to allow its implementation in concert with other, extant narrowband and conventional spreading technologies? I think not.
Rumor has it that enough of us HAVE requested ArcView and ArcIMS for Linux that they are going to be in Beta sometime Q2 or Q3. I'll see what I can do to confirm this next week...
HOWEVER, what OGC *is* working on, and in a fairly fractious and "open" manner, is standards for consistent handling of geospatial data. I can speak to some degree about datums, ellipsoids, geoids and coordinate transformations, and the infusion of new blood (Royal Dutch Sell have added a really talented geodesist) has been marvelous.
At least in this subcommittee, the intent is a consistent handling of transforms, something missing in the industry save among academics and folks like the National Geodetic Survey, where doing it "right" is more important than having a new "tool" to do it differently.
Further, the publication of GML-1 and the current efforts to make GML-2 a standard are significant events in facilitating datasharing. Real significant.
No, they're not dedicated to developing open source software. But the idea is open data standards so data can be distributed amongst the various packages represented, and some of the governmental agencies represented DO have open source in mind for their apps.
In my experience (non-trivial, actually) the current state of autonomous civil (L1, CA-code phase) positioning provides an accruace of about 6 meters 1d RMS (horizontal) and about 8m 1d RMS (vertical). Incorporation of a decent Differential GPS receiver and update system allows for improvements of about 1 order of magnitude (0.6m (H), 0.8m (V)).
The last time I checked, 60 cm was on the order of 24 inches... not 3 inches. Incorporation of WAAS (Wide Area Augmentation System) and eventually LAAS (Local Area Augmentation System) into thepicture will provide similar levels of improvement. but will NOT get you to the 1 cm level, reliably.
Geodetic positioning, where accuracies at or below the centimeter level are desired/required, require differencing the results from paired observations (2 receivers, one on a known monument) over a significant period of time (my current research, as well as that of the National Geodetic Survey (in work) have demonstrated observation times of 4 hours minimum, while I'm recommending a 50% increase in that duration while we're near the peak of the current sun spot cycle, because of the increased, and somewhat variable scintillation of the ionosphere.
Post-processing of the difference data result in a statistical answer for a baseline offset between the two sites. When incorporated with multiple observing sessions, a network of these baselines is developed, and the position of the point of interest is resolved with reference to the known, fixed and well-surveyed monuments, usually using the technique of least squares.
Where was I? Oh, yes... Autonomour hand-held concumer-grade receivers are generally good, now that Selective Availability has been removed, for accuracies on the order of 6-10 meters, with most degradation due to either the inadequacies of the receiver (increased processing noise, inadequate internal clock stability, poor algorithm implementation, failure to comply with IDC-GPS-200) or because of antenna anomalies. If you look at the error budgets for the GPS system, ignoring the problems associated with the now-defunct Selective Availability, a pessimist would conclude that your best hope for accuracy should be on the order of 29 meters (horizontal) for the Coarse Acquisition-Code solution, and 17 meters for the Precise Positioning Service code solution. That the results are better than that are a tribute to the system designers in the 1970s who dreamed this all up and started implementing it, as well as the current stable of designers and implementers (satellites, Military, and receiver manufacturers) who keep up with the system.
After their initial (disasterous) attempt to create a list presence, they seem to have gotten a better grip on dealing with the open-source community. Also, I've done some digging, and found myself more charitable to them when I learned more about them... both in the/. followup and from several friends in the Austin tech community. Their road is still rocky, so we may see these things on the surplus market sooner rather than later, but it might also be a Good Thing (C) for the open source community to see this concept of low-end entitlement grow some.
In fact, using something like the Optima line of batteries, sealed, depleted-electrolyte lead-acid, with a 30-psi seal, gas recombinant, with adequate thermal control to keep 'em from freezing and direct exposure to sunlight, should work. Careful definition of the power budget allows you to calculate how much battery you need; careful atention to over-sizing by some percentage precludes being overly surprised when something draws more current than you expected.
Add an orienting solar array and charging logic, steering diodes and such and you're in pretty good shape.
And, yes, I've got a clue of what the design parameters should be for something like this.
OK. I get in trouble with SOMEONE everytime I do this, so let me don the Kevlar...
1. There are 2 GPS bands, referred to as frequencies, centered at 1542.75 MHz (L1) and 1227.60 MHz (L2). Coarse Acquisition and P-code are available on L1 and P-(Y)-code on L2. Each band is 10 MHz wide, and the signals are categorized as direct sequence spread spectrum.
2. Coarse Acquisition (C/A) code was originally envisioned as a means of acquiring the Clock to allow for P-code acquisition. It was then noted that as an unclassified signal with the core navigation information, it could be used for civil applications with an accuracy degradation.
3. P-code was encrypted (became P(Y)-code about 7 years ago, but the exact date escapes me right now.
4. receivers with P-code capability are embargoed by the Munitions Export regulations. Go figure.
Standard Positioning Service utilizes L1 C/A code-phase, navigation and ephemeris data to determine a receiver's antenna position in a cartesian system that is rooughly Earth Centerd, Earth Fixed, and is based on the GRS80 ellipsoid. It uses ephemeris data (orbital parameters that can be propagated to a given time and used to calculate a satellite's instantaneous position) and navigation message data that amounts to precise timing information. The receiver's clock circuitry is phase-locked to a reference satellite's clock. The delay in signal timing between transmission by the satellite and reception at the receiver is correlated to the speed of light and a vector is established to each satellite in view. With _4_ satellites in the solution matrix, X, Y, and Z can be resolved in the matrix. If an altitude is known, it can be transformed to an appropriate offset value set (X,Y,Z) and 3 satellites can approximately establish the remaining parameters. Similarly, if TIME is already known, 3 satellites will allow calculation of all the remaining parameters at a given epoch, but really accurate clocks are harder to come by, than a reasonably disciplined quartz clock phase-locked to a satellite cesium or rhubidium standard. SPS provides an autonomous positional accuracy of approximately 29m RMS at 2 sigma and 56m vertical 2-sigma, RMS.
Precise positioning System uses a similar approach, but uses the P(Y)-code, which incorporates a higher chipping rate by an order of magnitude. Thus, the inherent accuracy at a given epoch is improved by a degree of granularity. PPS is stated to provide 14m 2d RMS horizontal accuracies autonomously and 33m vertically.
Engineers at Rockwell tell me that SpaceCom have been incorporating a few more bits of precision in the P-Code satellite ephemeris data, and that some receivers (Rockwell military hardware) now can routinely achieve 5m horizontal and 7m vertical accuracies autonomously.
As a rule of thumb, differential correction, via either real-time of post-processing data, affords a 10x improvement in accuracy providing all the games are appropriately played: The DGPS base is seeing all the same satellites as the rover, epochs are synchronized, there are no significant disturbances in troposphere or ionosphere between base and rover. Thus, SPS should see 2-3m 2d (hor) RMS accuracies with DGPS and 5-6m vertical accuracies, while PPS should see 0.5-1.5m 2d and 1-3m vertical accuracies. There are notable times when the error budget aligns in your favor and these accuracies improve.
Surveyors utilize the RF carrier wave characteristics to perform their magic, and also use the two seperate frequencies to calculate ionospheric propagation effects. Ionospheric modelling is not possible without at least 2 different frequencies, and diversity in spectrum accentuates the deltas. use of a reasonable model allows for a further reduction of the error budget.
In survey GPS observation, a baseline, or series of baselines, is eatablished between the point of interest and 'n' well-surveyed control monuments. Using the method of least-squares, a network of these points is adjusted for best error distuibution and the result ascertained. A satisfactory number of observations must be obtained, usually requiring observations at 30 sec epochs for 30 min to as long as 72 hours, depending on the application. A process known as integer ambiguity resolution may als be employed to establish on an epoch-by-epoch basis the integer number of carrier wave cycles between each satellite and the reciver; the fractional portion (phase angle) of the received signal is easy to measure. Resolving the integer ambiguity allows for a much finer estimate of satellite-to-receiver distance (vector magnitude) and thus a better solution of the position matrix.
If there are questions, follow up by e-mail and I'll either repost answers here, if there's a general call, or respond by private e-mail
For the record: the TerraServer imagery *is* from the Digital Ortho QuarterQuad program, not satellite imagery, save for a few demosntrations they had. The project with USGS ran out some time ago. USGS is not likely to pump money into it again, although you may see them setting up their OWN terraserver some day.
The DOQQ photography is dependent on the partnering and additional funding provided by the States. In Texas, for example, the DOQ imagery is in color IR, so land-use/land-classification work can proceed. The USGS standard base is black-and-white high-res. The standard altitude is 30,000 feet. Registration is done, these days, using GPS in a kinematic survey mode, which provides accuracies similar to (although theoretically better than) the older registration marks on the ground. Realize also that
It currently takes 4 years to fly, process, register, rectify, catalog, archive and release the State of Texas' DOQQs. The State (http://www.tnris.state.texas.us/DigitalData/dows. htm) has spent a lot of time, effort and money doing all this, and placing the most current datasets in the public domain.
The terraserver project was a joint venture between Micro$oft and the US Geological Survey to see if they _could_ serve that much imagery. While its speed tends to make it less than an overwhelming success, the fact that thec usually get the right image to you is a real hallmark. gerry
Having worked at NASA, I took exception to Sensenbrenner's comments toward NASA when John Glenn was announced for the Shuttle mission. Rep. Sensenbrenner basically said NASA had assigned 2 cardiovascular specialists to babysit Glenn.
My e-mail shot got a stock e-mail response, followed with a stock postal response saying they didn't bother with anyone outside Sensenbrenner's district. When I fired a truly irate response back, suggesting a call to the media, I got a letter from the Representative, or someone who at least took the time to apologize in his name and admit the remarks were out of line, in 2 days. By snail mail.
The consideration that e-mail is too convenient, is specious. It's the way I do business now. It *is* convenient. It is easier for me to e-mail my congressman than to remember how to get the Laserjet to address an envelope.
Let's get Congress to start considering our input. It's useful to listen to the tech-savvy folks sometimes...
Standardized and specialized tests suggest that I *AM* a genius. Pico doesn't break class structures like VC++. I *AM* a programmer whether I install RPMs on Linux or not. The command line and the options give me more utility than a pull-down menu.
I have no social life to speak of, save teaching, guest-lecturing to professional societies, presentations at meetings, and Cub Scouts... I have a wife and 3 kids. Most of my athletic ability (long-distance cycling, rock climbing) ended 2 years ago with a nasty work-related injury to both legs. But I'm coming back, and hiking's available again.
At least I can identify the difference between an operating system and a kludgy monitor with a badly implemented GUI.
Harris Semiconductor has spent a lot of time and effort making the iAPx86 processor line radiation hardened. And it works. But it doesn't work nearly as well for the newer generation as the older.
The tighter dies are more sensitive to charged particle events (Single Event Upsets, 'SEUs'), and the thinner nature of deposition makes the hardware more prone to heavy particle (permanent) events.
Older CMOS processors and memory were less prone to both types of error because of line thickness and spacing, depth of deposition and because of the voltage levels used.
The COSMAC 1802 is still a widely used space-borne processor!
Rad-hard planning is a non-trivial task, expecially when it's new to the designer. Trying to develop a design for a medium-altitude satellite was a challenge that a group I was once involved with was never able to satisfactorily overcome, and we lost the project because of it.
Having watched the evolution of the Shuttle system General Purpose Computers (GPCs) migrate from IBM proprietary processors to 80186s was a painful process. The software was clean-roomed, and 2 seperate groups of people developed code with identical functionality, while a 3rd group was responsible for testing. None of the groups was supposed to interact at all, and to the best of my knowledge, didn't even talk over beer... I suspect the Hubble ports are similar, although the rules for Manned Spaceflight code are more stringent.
Nope. He didn't have to do something wrong, nor you, necessarily something either right or wrong, to achieve your individual results. You merely discovered teh PRN generator in the Win install code. Periodically, it does what it actually claims to do, rather than obliterating other code and software, and mortally injuring itself on first installs.
Checked lately? I think you'll find there's now a vector component to GRASS.
And, ESRI is thumbing their noses at Linux as a passing fad... or so they've told me. Matter of fact, so is ERDAS.
Research Systems may well give ERDAS a run for their money: I've had real good results lately with their IDL/ENVI suite for remote sensing, even if I can't get AVHRR data to load.
As things stand now, implementing a GIS is non-trivial. You've got to be well-trained in the software, knowledgable in databases, and have a clue about geodesy. Without these, and other key components, you have a program that ends up making maps which are then of questionable utility.
I firmly agree with paragraph 2, however. I foresee the day when the "GIS Industry" fades away save for the academics teaching it, and they're going to fade to obscurity only a bit more slowly. Spatial representation of data is a fairly natural method of analysis and display. That it's not been widely implemented so far has, as much as anything, been the result of database limitations... and more in theory than applicaiton implementations. Indeed, I'm willing to bet that within 10 years, perhaps less, the "big names" in GIS software in the industrial markey will be looking at the Microsofts, Corels, Applixes, and other suite makers, and wondering how the market got there.
As for datasets: yes, our money paid for their development. A lot of them are available for free, or the cost of duplication. You've got to know where to get them... In Texas, the Texas Natural Resources Information Service (TNRIS: http://www.tnris.state.tx.us) has a lot of geodata for the whole state available for download... for free. Storage limitations preclude putting ALL of the data out that way, but they're working on that, and the cost of distribution for non-downloadable data is restricted by state law to the costs of duplication, media and delivery.
Landsat-7 data are available from USGS now for about 1/2 the cost of scenes from previous birds. USGS is developing a warehouse of older SPOT-Image data that they're trying to make available for duplication costs only.
SO: There's a lot out on that front. You have to look, I have to scout out tidbits at the meetings. Now: Can someone tell me why the DoD National Imagery and Mapping Agency's Level 1 (10m) Digital Terrain Elevation Models (DTEDs) are classified for the Continental US? Because of that, I can't get the new NGS densifications of height vs. local gravity in Geoid99!
I was truly fascinated to read the page and their misrepresentations, representations, and FUD yesterday, while waiting for a compile to finish. Went home and did a little channel surfing, and one of their customers, Nasdaq, was in the news for the software and OS. Seems they installed some new software and it's so efficient that it's slowed their processing down to a crawl. Brokerage firms and traders big and small were howling and calling for SEC investigations and lawsuits.
High resolution imagery is used for mission planning. Multiple images, offset spatially and appropriately orthorectified, with good ephemeris data, yield information about topography. You know how the terrain looks, where the linear features are, what their dimensions are, where the likely security checkpoints and rallying points are. Simply put, information is the key to a successful operation if you're trying to get in, or get a package in.
Re:You mean "protested the INVASION of Vietnam"...
on
U.S. Army Testing Jini
·
· Score: 1
First, the conflict in Viet Nam was technically a "police action" similar to that in Korea. 2. The US was asked for technical and military air by a legally constituted government in S. Viet Nam. We shall leave for the time being the question of corruption and whether the US intelligence Community supported this government beyond its useful life. 3. The escalation of the conflict to include US fighting troops (beyond the initial "advisor" role was based on an invitation to assist from the the S. Viet Nam government. 4. In virtually all logical accounts, the North Vietnamese regular army and the Viet Cong lost the conflict militarily. 5. The failure of the US military and its allies in the Viet Nam conflict to "win" overall was almost completely due to the strong political input to the selection of targets and campaigns. Tactical target selection from Washington is a poor way to run a war. Further, denying warfighters the opportunity to hit known strategically important targets (we aren't talking nukes, kiddies, we're talking large-scale impact on warfighting) such as anti-air assets, harbor wharfs and load-handling facilities, railroad assets communicating with principal suppliers of war materiale, etc., has profound impacts on morale and discipline. The final loss of the country to the North Vietnamese was a result of a number of bad political decisions forced upon a warfighting military capable (and proven so) of winning that conflict. The apparent loss was highlighted in the popular press, but never told the real stories of overwhelming military successes... body counts don't tell it all, but ability to mobilize coherent fighting forces do, and the VC and NVA were unable to do that in real terms by the time we withdrew... in disarray. One significant thing to come out of this debacle was an effective remaking of the US military structure. Officers who saw the problems inflicted upon the US military sought to prevent that from happening agian. Operations in Panama, the 1992 Desert Shield/Desert Storm operation and in Kosovo have shown that by designating political goals and objectives, then keeping the political hacks out of the direct planning, missions can be carried through. Sorry if your facts are in error.
I'll have to snag a copy today and review the article. I'm not up to speed on fractal antenna design, so it'll be interesting. Remember that, however, for GPS antennas, reception of multiple-reflectionsignals degrades your result rather than improving it. You want to maintain a Right Hand Circular polarization to diminish multipath-delayed signals (13-20dB improvement in signal selection; less confused receiver) for your best GPS performance.
And, yes, I realize that it'd be nice toget better performance under trees and in urban canyons but that's more a function of the GPS satellties transmitting a paltry amount of power relative to free-space loss and range. But a lot of power when one considers power generation on a satellite!
Slashdot Mirror
"had the impression (uh, no quotes, sorry) that one of the primary problems was not simply pumping blood, but pumping the right amount of blood.
>A natural heart adjusts depending on the needs of the body - for an artificial heart, is this a big issue? Does anyone know how it is addressed, in general, and in this specific case?"
One of several approaches to this is to use a pneumatic "volume compensation device to allow some variability in the area behind the pump for filling. The body has a series of checks and balances, but for flow regulation it's mostly based on venous return: The arteries can constrict causing the blood pressure to rise in regional or overall (systemic) fashion, but there is precious little that will cause the veins to constrict. While unanticipated venous dilation would cause problems, generally, if the physiology is, overall, intact, this will _not_ happen.
So, what I'm suggesting here is that the flow could well be regulated in several ways. I will have to do some research on the Abiomed device in its current state, but the pump could well have a central aortic pressure sensor and attempt to pump more blood (ie., faster) to maintain a mean pressure; the pump could be triggered to "fire" when a certain fill point in the ventricular chambers is reached, or the pump could be set to a fixed rate and simply pump the available blood at the time it's supposed to pump, little or lots, depending on venous return.
Having been out of the game for a while I'm not sure anymore which of the methods is prevalent, but the first 2, in isolation or combination are physiologically attractive.
I spent a number of years in the articifial heart field while at Texas Heart Institute. It really *IS* big news.
All the artificial heart implants in humans have been tethered implants. The patient has had to be tied to a console or power source, electrical or air, with no more than brief respites of untethered activity. The devices have been harsh on quality of life, and a whole host of physiologic functions. They have not allowed normal interaction with other humans, and those few patients who've ventured outside the walls of the hospital were making brief visits away, not returning to the world.
The Abiomed pump is small enough to reasonably be implanted, and reliable enough to expect it to work well. The centers selected for the initial implants have sufficient experience with animal implantation, AND various human procedures of a more mundane variety, to expect them to be able to manage the patients well, indeed.
We're about 9 years behind where I thought we'd get to with a really viable, implantable heart, mainly because of the costs necessary to support this sort of research. It's long overdue.
I suspect that the 125,000 potential patients Abiomed cited in the article may be an understatement. Doesn't matter. If this allows some patients who were dying waiting for a transplant...or who were deemed not good candidates for the scarce resource of a donated heart... a shot at a good quality life and a time extension, this is WONDERFUL news.
Makes me wish I'd stayed in the game.
My step-daughter's now 16, and a bit more "worldly" than I'd like to admit. But she's had net access for the last 5 years, pretty much freely, and virtually unfettered access for the last 3.5 years.
... and we nailed him, too.
I do networking for a living. I've periodically dropped a sniffer on her stream and captured content to see what's going on. My wife and I have discreetly used the garnered info, when we've felt we had to, to forestall some potentially hairy problems.
I *don't* want to admit that she's going to have sex as a teenager. But, you know, overall she's a pretty sharp kid, and we do NOT monitor her activity on the Web tightly. We have not had the need to do so for some time: her judgement and her decisions to come to me with 'Net-related questions and problems have indicated she is pretty safe on the Web all by herself.
We had the "It's a mean Net out there" talk back when she was first starting, and again about 18 months ago. She knows she can always come to me with stuff regarding the Net and the Web. And she has, even coming for advice when a friend of hers started having someone spoof her mail to friends in a particularly offensive way
I tend to agree: Trust and time are the keys. And while I've monitored, I've also trusted... Consider the monitoring a form of verification, and realize that we take our role in shaping her life pretty seriously.
g.
Actually, the NTIA assessment of this stuff is that, Yes, it may work, and, based on actual and theoretical evaluations, its potential for harmful interference is significant above certain threshholds for power and "chipping rate."
The ultrawideband nature of this approach appears to offer a splendid opportunity to raise the noise floor across a large region of RF spectrum, to the point that little things, like GPS navigation and cell phones as we know them today will be adversely affected.
You're right: It's a neat development. But has it been appropriately evaluated to allow its implementation in concert with other, extant narrowband and conventional spreading technologies? I think not.
Rumor has it that enough of us HAVE requested ArcView and ArcIMS for Linux that they are going to be in Beta sometime Q2 or Q3. I'll see what I can do to confirm this next week...
HOWEVER, what OGC *is* working on, and in a fairly fractious and "open" manner, is standards for consistent handling of geospatial data. I can speak to some degree about datums, ellipsoids, geoids and coordinate transformations, and the infusion of new blood (Royal Dutch Sell have added a really talented geodesist) has been marvelous.
At least in this subcommittee, the intent is a consistent handling of transforms, something missing in the industry save among academics and folks like the National Geodetic Survey, where doing it "right" is more important than having a new "tool" to do it differently.
Further, the publication of GML-1 and the current efforts to make GML-2 a standard are significant events in facilitating datasharing. Real significant.
No, they're not dedicated to developing open source software. But the idea is open data standards so data can be distributed amongst the various packages represented, and some of the governmental agencies represented DO have open source in mind for their apps.
--gerry
Debunking time, in general terms.
In my experience (non-trivial, actually) the current state of autonomous civil (L1, CA-code phase) positioning provides an accruace of about 6 meters 1d RMS (horizontal) and about 8m 1d RMS (vertical). Incorporation of a decent Differential GPS receiver and update system allows for improvements of about 1 order of magnitude (0.6m (H), 0.8m (V)).
The last time I checked, 60 cm was on the order of 24 inches... not 3 inches. Incorporation of WAAS (Wide Area Augmentation System) and eventually LAAS (Local Area Augmentation System) into thepicture will provide similar levels of improvement. but will NOT get you to the 1 cm level, reliably.
Geodetic positioning, where accuracies at or below the centimeter level are desired/required, require differencing the results from paired observations (2 receivers, one on a known monument) over a significant period of time (my current research, as well as that of the National Geodetic Survey (in work) have demonstrated observation times of 4 hours minimum, while I'm recommending a 50% increase in that duration while we're near the peak of the current sun spot cycle, because of the increased, and somewhat variable scintillation of the ionosphere.
Post-processing of the difference data result in a statistical answer for a baseline offset between the two sites. When incorporated with multiple observing sessions, a network of these baselines is developed, and the position of the point of interest is resolved with reference to the known, fixed and well-surveyed monuments, usually using the technique of least squares.
Where was I? Oh, yes... Autonomour hand-held concumer-grade receivers are generally good, now that Selective Availability has been removed, for accuracies on the order of 6-10 meters, with most degradation due to either the inadequacies of the receiver (increased processing noise, inadequate internal clock stability, poor algorithm implementation, failure to comply with IDC-GPS-200) or because of antenna anomalies. If you look at the error budgets for the GPS system, ignoring the problems associated with the now-defunct Selective Availability, a pessimist would conclude that your best hope for accuracy should be on the order of 29 meters (horizontal) for the Coarse Acquisition-Code solution, and 17 meters for the Precise Positioning Service code solution. That the results are better than that are a tribute to the system designers in the 1970s who dreamed this all up and started implementing it, as well as the current stable of designers and implementers (satellites, Military, and receiver manufacturers) who keep up with the system.
After their initial (disasterous) attempt to create a list presence, they seem to have gotten a better grip on dealing with the open-source community. Also, I've done some digging, and found myself more charitable to them when I learned more about them... both in the /. followup and from several friends in the Austin tech community. Their road is still rocky, so we may see these things on the surplus market sooner rather than later, but it might also be a Good Thing (C) for the open source community to see this concept of low-end entitlement grow some.
In fact, using something like the Optima line of batteries, sealed, depleted-electrolyte lead-acid, with a 30-psi seal, gas recombinant, with adequate thermal control to keep 'em from freezing and direct exposure to sunlight, should work. Careful definition of the power budget allows you to calculate how much battery you need; careful atention to over-sizing by some percentage precludes being overly surprised when something draws more current than you expected.
Add an orienting solar array and charging logic, steering diodes and such and you're in pretty good shape.
And, yes, I've got a clue of what the design parameters should be for something like this.
OK. I get in trouble with SOMEONE everytime I do this, so let me don the Kevlar...
1. There are 2 GPS bands, referred to as frequencies, centered at 1542.75 MHz (L1) and 1227.60 MHz (L2). Coarse Acquisition and P-code are available on L1 and P-(Y)-code on L2. Each band is 10 MHz wide, and the signals are categorized as direct sequence spread spectrum.
2. Coarse Acquisition (C/A) code was originally envisioned as a means of acquiring the Clock to allow for P-code acquisition. It was then noted that as an unclassified signal with the core navigation information, it could be used for civil applications with an accuracy degradation.
3. P-code was encrypted (became P(Y)-code about 7 years ago, but the exact date escapes me right now.
4. receivers with P-code capability are embargoed by the Munitions Export regulations. Go figure.
Standard Positioning Service utilizes L1 C/A code-phase, navigation and ephemeris data to determine a receiver's antenna position in a cartesian system that is rooughly Earth Centerd, Earth Fixed, and is based on the GRS80 ellipsoid. It uses ephemeris data (orbital parameters that can be propagated to a given time and used to calculate a satellite's instantaneous position) and navigation message data that amounts to precise timing information. The receiver's clock circuitry is phase-locked to a reference satellite's clock. The delay in signal timing between transmission by the satellite and reception at the receiver is correlated to the speed of light and a vector is established to each satellite in view. With _4_ satellites in the solution matrix, X, Y, and Z can be resolved in the matrix. If an altitude is known, it can be transformed to an appropriate offset value set (X,Y,Z) and 3 satellites can approximately establish the remaining parameters. Similarly, if TIME is already known, 3 satellites will allow calculation of all the remaining parameters at a given epoch, but really accurate clocks are harder to come by, than a reasonably disciplined quartz clock phase-locked to a satellite cesium or rhubidium standard. SPS provides an autonomous positional accuracy of approximately 29m RMS at 2 sigma and 56m vertical 2-sigma, RMS.
Precise positioning System uses a similar approach, but uses the P(Y)-code, which incorporates a higher chipping rate by an order of magnitude. Thus, the inherent accuracy at a given epoch is improved by a degree of granularity. PPS is stated to provide 14m 2d RMS horizontal accuracies autonomously and 33m vertically.
Engineers at Rockwell tell me that SpaceCom have been incorporating a few more bits of precision in the P-Code satellite ephemeris data, and that some receivers (Rockwell military hardware) now can routinely achieve 5m horizontal and 7m vertical accuracies autonomously.
As a rule of thumb, differential correction, via either real-time of post-processing data, affords a 10x improvement in accuracy providing all the games are appropriately played: The DGPS base is seeing all the same satellites as the rover, epochs are synchronized, there are no significant disturbances in troposphere or ionosphere between base and rover. Thus, SPS should see 2-3m 2d (hor) RMS accuracies with DGPS and 5-6m vertical accuracies, while PPS should see 0.5-1.5m 2d and 1-3m vertical accuracies. There are notable times when the error budget aligns in your favor and these accuracies improve.
Surveyors utilize the RF carrier wave characteristics to perform their magic, and also use the two seperate frequencies to calculate ionospheric propagation effects. Ionospheric modelling is not possible without at least 2 different frequencies, and diversity in spectrum accentuates the deltas. use of a reasonable model allows for a further reduction of the error budget.
In survey GPS observation, a baseline, or series of baselines, is eatablished between the point of interest and 'n' well-surveyed control monuments. Using the method of least-squares, a network of these points is adjusted for best error distuibution and the result ascertained. A satisfactory number of observations must be obtained, usually requiring observations at 30 sec epochs for 30 min to as long as 72 hours, depending on the application. A process known as integer ambiguity resolution may als be employed to establish on an epoch-by-epoch basis the integer number of carrier wave cycles between each satellite and the reciver; the fractional portion (phase angle) of the received signal is easy to measure. Resolving the integer ambiguity allows for a much finer estimate of satellite-to-receiver distance (vector magnitude) and thus a better solution of the position matrix.
If there are questions, follow up by e-mail and I'll either repost answers here, if there's a general call, or respond by private e-mail
Regards,
Gerry Creager
Texas A&M University
For the record: the TerraServer imagery *is* from the Digital Ortho QuarterQuad program, not satellite imagery, save for a few demosntrations they had. The project with USGS ran out some time ago. USGS is not likely to pump money into it again, although you may see them setting up their OWN terraserver some day.
. htm) has spent a lot of time, effort and money doing all this, and placing the most current datasets in the public domain.
The DOQQ photography is dependent on the partnering and additional funding provided by the States. In Texas, for example, the DOQ imagery is in color IR, so land-use/land-classification work can proceed. The USGS standard base is black-and-white high-res. The standard altitude is 30,000 feet. Registration is done, these days, using GPS in a kinematic survey mode, which provides accuracies similar to (although theoretically better than) the older registration marks on the ground. Realize also that
It currently takes 4 years to fly, process, register, rectify, catalog, archive and release the State of Texas' DOQQs. The State (http://www.tnris.state.texas.us/DigitalData/dows
The terraserver project was a joint venture between Micro$oft and the US Geological Survey to see if they _could_ serve that much imagery. While its speed tends to make it less than an overwhelming success, the fact that thec usually get the right image to you is a real hallmark.
gerry
Having worked at NASA, I took exception to Sensenbrenner's comments toward NASA when John Glenn was announced for the Shuttle mission. Rep. Sensenbrenner basically said NASA had assigned 2 cardiovascular specialists to babysit Glenn.
My e-mail shot got a stock e-mail response, followed with a stock postal response saying they didn't bother with anyone outside Sensenbrenner's district. When I fired a truly irate response back, suggesting a call to the media, I got a letter from the Representative, or someone who at least took the time to apologize in his name and admit the remarks were out of line, in 2 days. By snail mail.
The consideration that e-mail is too convenient, is specious. It's the way I do business now. It *is* convenient. It is easier for me to e-mail my congressman than to remember how to get the Laserjet to address an envelope.
Let's get Congress to start considering our input. It's useful to listen to the tech-savvy folks sometimes...
Standardized and specialized tests suggest that I *AM* a genius. Pico doesn't break class structures like VC++. I *AM* a programmer whether I install RPMs on Linux or not. The command line and the options give me more utility than a pull-down menu.
I have no social life to speak of, save teaching, guest-lecturing to professional societies, presentations at meetings, and Cub Scouts... I have a wife and 3 kids. Most of my athletic ability (long-distance cycling, rock climbing) ended 2 years ago with a nasty work-related injury to both legs. But I'm coming back, and hiking's available again.
At least I can identify the difference between an operating system and a kludgy monitor with a badly implemented GUI.
Harris Semiconductor has spent a lot of time and effort making the iAPx86 processor line radiation hardened. And it works. But it doesn't work nearly as well for the newer generation as the older.
The tighter dies are more sensitive to charged particle events (Single Event Upsets, 'SEUs'), and the thinner nature of deposition makes the hardware more prone to heavy particle (permanent) events.
Older CMOS processors and memory were less prone to both types of error because of line thickness and spacing, depth of deposition and because of the voltage levels used.
The COSMAC 1802 is still a widely used space-borne processor!
Rad-hard planning is a non-trivial task, expecially when it's new to the designer. Trying to develop a design for a medium-altitude satellite was a challenge that a group I was once involved with was never able to satisfactorily overcome, and we lost the project because of it.
Having watched the evolution of the Shuttle system General Purpose Computers (GPCs) migrate from IBM proprietary processors to 80186s was a painful process. The software was clean-roomed, and 2 seperate groups of people developed code with identical functionality, while a 3rd group was responsible for testing. None of the groups was supposed to interact at all, and to the best of my knowledge, didn't even talk over beer... I suspect the Hubble ports are similar, although the rules for Manned Spaceflight code are more stringent.
Nope. He didn't have to do something wrong, nor you, necessarily something either right or wrong, to achieve your individual results. You merely discovered teh PRN generator in the Win install code. Periodically, it does what it actually claims to do, rather than obliterating other code and software, and mortally injuring itself on first installs.
Checked lately? I think you'll find there's now a vector component to GRASS.
And, ESRI is thumbing their noses at Linux as a passing fad... or so they've told me. Matter of fact, so is ERDAS.
Research Systems may well give ERDAS a run for their money: I've had real good results lately with their IDL/ENVI suite for remote sensing, even if I can't get AVHRR data to load.
As things stand now, implementing a GIS is non-trivial. You've got to be well-trained in the software, knowledgable in databases, and have a clue about geodesy. Without these, and other key components, you have a program that ends up making maps which are then of questionable utility.
I firmly agree with paragraph 2, however. I foresee the day when the "GIS Industry" fades away save for the academics teaching it, and they're going to fade to obscurity only a bit more slowly. Spatial representation of data is a fairly natural method of analysis and display. That it's not been widely implemented so far has, as much as anything, been the result of database limitations... and more in theory than applicaiton implementations. Indeed, I'm willing to bet that within 10 years, perhaps less, the "big names" in GIS software in the industrial markey will be looking at the Microsofts, Corels, Applixes, and other suite makers, and wondering how the market got there.
As for datasets: yes, our money paid for their development. A lot of them are available for free, or the cost of duplication. You've got to know where to get them... In Texas, the Texas Natural Resources Information Service (TNRIS: http://www.tnris.state.tx.us) has a lot of geodata for the whole state available for download... for free. Storage limitations preclude putting ALL of the data out that way, but they're working on that, and the cost of distribution for non-downloadable data is restricted by state law to the costs of duplication, media and delivery.
Landsat-7 data are available from USGS now for about 1/2 the cost of scenes from previous birds. USGS is developing a warehouse of older SPOT-Image data that they're trying to make available for duplication costs only.
SO: There's a lot out on that front. You have to look, I have to scout out tidbits at the meetings. Now: Can someone tell me why the DoD National Imagery and Mapping Agency's Level 1 (10m) Digital Terrain Elevation Models (DTEDs) are classified for the Continental US? Because of that, I can't get the new NGS densifications of height vs. local gravity in Geoid99!
I was truly fascinated to read the page and their misrepresentations, representations, and FUD yesterday, while waiting for a compile to finish. Went home and did a little channel surfing, and one of their customers, Nasdaq, was in the news for the software and OS. Seems they installed some new software and it's so efficient that it's slowed their processing down to a crawl. Brokerage firms and traders big and small were howling and calling for SEC investigations and lawsuits.
THAT's a testimonial for you!
High resolution imagery is used for mission planning. Multiple images, offset spatially and appropriately orthorectified, with good ephemeris data, yield information about topography. You know how the terrain looks, where the linear features are, what their dimensions are, where the likely security checkpoints and rallying points are. Simply put, information is the key to a successful operation if you're trying to get in, or get a package in.
First, the conflict in Viet Nam was technically a "police action" similar to that in Korea. 2. The US was asked for technical and military air by a legally constituted government in S. Viet Nam. We shall leave for the time being the question of corruption and whether the US intelligence Community supported this government beyond its useful life. 3. The escalation of the conflict to include US fighting troops (beyond the initial "advisor" role was based on an invitation to assist from the the S. Viet Nam government. 4. In virtually all logical accounts, the North Vietnamese regular army and the Viet Cong lost the conflict militarily. 5. The failure of the US military and its allies in the Viet Nam conflict to "win" overall was almost completely due to the strong political input to the selection of targets and campaigns. Tactical target selection from Washington is a poor way to run a war. Further, denying warfighters the opportunity to hit known strategically important targets (we aren't talking nukes, kiddies, we're talking large-scale impact on warfighting) such as anti-air assets, harbor wharfs and load-handling facilities, railroad assets communicating with principal suppliers of war materiale, etc., has profound impacts on morale and discipline. The final loss of the country to the North Vietnamese was a result of a number of bad political decisions forced upon a warfighting military capable (and proven so) of winning that conflict. The apparent loss was highlighted in the popular press, but never told the real stories of overwhelming military successes... body counts don't tell it all, but ability to mobilize coherent fighting forces do, and the VC and NVA were unable to do that in real terms by the time we withdrew... in disarray. One significant thing to come out of this debacle was an effective remaking of the US military structure. Officers who saw the problems inflicted upon the US military sought to prevent that from happening agian. Operations in Panama, the 1992 Desert Shield/Desert Storm operation and in Kosovo have shown that by designating political goals and objectives, then keeping the political hacks out of the direct planning, missions can be carried through. Sorry if your facts are in error.
I'll have to snag a copy today and review the article. I'm not up to speed on fractal antenna design, so it'll be interesting. Remember that, however, for GPS antennas, reception of multiple-reflectionsignals degrades your result rather than improving it. You want to maintain a Right Hand Circular polarization to diminish multipath-delayed signals (13-20dB improvement in signal selection; less confused receiver) for your best GPS performance.
And, yes, I realize that it'd be nice toget better performance under trees and in urban canyons but that's more a function of the GPS satellties transmitting a paltry amount of power relative to free-space loss and range. But a lot of power when one considers power generation on a satellite!