|
Low-Pass Filter posted:We've played with Terrastar subscriptions and can get about 20cm with motion with a single receiver and L1/L2. RTK gps can easily get under 10cm with person level movement; this is how surveyors lay out roads and stuff accurately. Still need somewhat expensive L1/L2/L5 antennas tho. what if you're moving next to a tall building or underneath some trees how much does that affect the accuracy
|
# ? Feb 4, 2019 19:04 |
|
|
# ? Mar 29, 2024 14:12 |
|
If I remember correctly all GPS based navigation systems for mobile applications blend the GPS feed with inertial navigation for continuous tracking, but LPF is clearly the authority on the matter.
|
# ? Feb 4, 2019 19:29 |
|
COMPAGNIE TOMMY posted:oh yes, I see cyclists "combining" cycling with public transportation by clogging the diamond lane and preventing the bus from passing/moving faster than a bike in first gear thats the top speed of the bus anyway cause its gotta stop in 10 feet
|
# ? Feb 4, 2019 19:36 |
|
Captain Foo posted:driving sucks rear end and i wish i didn't have to do it so much evil_bunnY posted:Driving outside of spirited twisty roads or a track loving blows, my dude Combat Theory posted:A cool and good position iospace posted:My dude, have you heard of bicycles and public transit? You can even combine the two! trap sprung
|
# ? Feb 4, 2019 19:38 |
|
Combat Theory posted:If I remember correctly all GPS based navigation systems for mobile applications blend the GPS feed with inertial navigation for continuous tracking, but LPF is clearly the authority on the matter. yeah without dead reckoning gps sucks for navigation.
|
# ? Feb 4, 2019 19:40 |
|
Combat Theory posted:If I remember correctly all GPS based navigation systems for mobile applications blend the GPS feed with inertial navigation for continuous tracking, but LPF is clearly the authority on the matter. phones at least also use cell towers and other information. https://en.wikipedia.org/wiki/Assisted_GPS
|
# ? Feb 4, 2019 19:52 |
|
these ideas are too complicated just use a self-driving horse. strap it into AR blinders that show a picture of a carrot along the chosen route. give away the system for free and make a killing with subscription horse poop removal services. recycle the poop into bricks for sustainable concrete. bam ill take my VC money now
|
# ? Feb 4, 2019 19:56 |
|
jet airliers use a highly complicated system of gyros, inertia tracking, magnetic compass and probably other poo poo I forgot about for navigation and they'll still drift a few dozen miles per flight, so inertial navigation would be kinda pointless for cars where the difference of 1m is the difference between road and ditch.
|
# ? Feb 4, 2019 20:18 |
|
Truga posted:jet airliers use a highly complicated system of gyros, inertia tracking, magnetic compass and probably other poo poo I forgot about for navigation and they'll still drift a few dozen miles per flight, so inertial navigation would be kinda pointless for cars where the difference of 1m is the difference between road and ditch. i mean airliners are also going much faster and dealing with stuff like turbulence, and the error is cumulative over distance so the difference between going 1000 miles and going 10 miles is pretty significant, but regardless putting an IMU in a car is still probably pointless
|
# ? Feb 4, 2019 20:28 |
|
Shame Boy posted:i mean airliners are also going much faster and dealing with stuff like turbulence, and the error is cumulative over distance so the difference between going 1000 miles and going 10 miles is pretty significant, but regardless putting an IMU in a car is still probably pointless those also cost about what an entire car does
|
# ? Feb 4, 2019 20:39 |
|
Shame Boy posted:i mean airliners are also going much faster and dealing with stuff like turbulence, and the error is cumulative over distance so the difference between going 1000 miles and going 10 miles is pretty significant, but regardless putting an IMU in a car is still probably pointless I mean we gotta diffentiate a bit here. Every car with an electronic stability program already has a full 2 or 3 axis IMU on board, it's just some orders of magnitude cheaper and less accurate than what's used for long range inertial navigation. So has every Tom Tom or other windscreen Navi. Our Uni wants to Lauch a self driving car competition next year, I'll keep an eye on it but I expect little apart of hot air from it, given that the general attitude seems to be "we have no idea about self driving cars which is good because it means we won't get gated into common thinking approaches" I hear a little Elon fanboy fapping in the distant corner of the campus. Combat Theory fucked around with this message at 20:48 on Feb 4, 2019 |
# ? Feb 4, 2019 20:44 |
|
Combat Theory posted:I mean we gotta diffentiate a bit here. Every car with an electronic stability program already has a full 2 or 3 axis IMU on board, it's just some orders of magnitude cheaper and less accurate than what's used for long range inertial navigation. So has every Tom Tom or other windscreen Navi. "some" is doing a lot of work for this, a good laser ring gyro that is part of an imu system is $30k. a good automotive grade one is $5 from digikey which probably means $2 direct https://www.digikey.com/product-detail/en/stmicroelectronics/AIS3624DQTR/497-16268-2-ND/5402876
|
# ? Feb 4, 2019 20:47 |
|
I have read about laser ring gyros and I think my abstraction ability has reached its limits for the time being.
|
# ? Feb 4, 2019 20:50 |
|
Shame Boy posted:huh ok, though i thought surveyors like, tended to stay in one place when actually taking a measurement
|
# ? Feb 4, 2019 20:54 |
|
just have a fire truck drive where you want to go in front of you, duh
|
# ? Feb 4, 2019 20:57 |
|
Combat Theory posted:I have read about laser ring gyros and I think my abstraction ability has reached its limits for the time being.
|
# ? Feb 4, 2019 20:57 |
|
pretty much every fancy laser measurement thing is "lasers are coherent light so you can measure minute changes to its properties and know its due to some outside factor" so like, changes in a speckle pattern can be used to measure bloodflow or measure depth or all sorts of other stuff depending on what you build the main problem is that you're usually looking for minute changes so that means $$$$ for the optics
|
# ? Feb 4, 2019 21:15 |
|
should Musk be going on (what I assume is) some youtube meme show? is he even allowed by his sitters to do dumb stunts like this after the weed thing and every other thing?
|
# ? Feb 4, 2019 23:32 |
|
he keeps yes men around so yeah he's going to do whatever he wants.
|
# ? Feb 5, 2019 00:04 |
|
hobbesmaster posted:pretty much every fancy laser measurement thing is "lasers are coherent light so you can measure minute changes to its properties and know its due to some outside factor" see: fiber Bragg grating sensors. they’re really cool and can be used to measure an astonishing number of physical things you wouldn’t expect
|
# ? Feb 5, 2019 00:04 |
|
PleasureKevin posted:should Musk be going on (what I assume is) some youtube meme show? is he even allowed by his sitters to do dumb stunts like this after the weed thing and every other thing? (he claims that) he gets to decide whether or not something gets run by his sitters, iirc
|
# ? Feb 5, 2019 00:07 |
|
Shaggar posted:thats the top speed of the bus anyway cause its gotta stop in 10 feet
|
# ? Feb 5, 2019 00:08 |
|
Shame Boy posted:i mean airliners are also going much faster and dealing with stuff like turbulence, and the error is cumulative over distance so the difference between going 1000 miles and going 10 miles is pretty significant, but regardless putting an IMU in a car is still probably pointless love 2 spend 20 minutes parked in the driveway spinning up the gyros and aligning my inertial platforms every time i wanna go for a drive
|
# ? Feb 5, 2019 01:18 |
|
same but getting pork gyros and aligning them with my mouth
|
# ? Feb 5, 2019 01:22 |
|
Shame Boy posted:huh ok, though i thought surveyors like, tended to stay in one place when actually taking a measurement Well, the RTK base station does stay in one place and has to converge for like 20-30 minutes. Then the rover (i.e. surveyor guy with rod) can move around p much all he wants. If you're trying to locate, say, all the points on a foundation, or locate a water boundary, you're going to collect a lot of points, and you're not going to occupy them for very long. You can just walk up with the rover and hit "collect," no real need to wait around. If you're close enough to a COORS station, and are able to log raw carrier phase info, you can post process and get sub meter accuracy, but thats, uh, not very useful for a car. The equipment for this is like 10-20k tho. Low-Pass Filter fucked around with this message at 01:58 on Feb 5, 2019 |
# ? Feb 5, 2019 01:42 |
|
Sagebrush posted:love 2 spend 20 minutes parked in the driveway spinning up the gyros and aligning my inertial platforms every time i wanna go for a drive remember autonomous cars will always be on and roaming the street looking for fares
|
# ? Feb 5, 2019 01:55 |
|
H.P. Hovercraft posted:what if you're moving next to a tall building or underneath some trees how much does that affect the accuracy We never flew any UAV's under tree canopy or near tall buildings lol, so I can't give solid numbers. I'll ask my ex-surveyor co-worker if I can get some numbers. Theres two components in this environment that affect the accuracy. One is Multipath, where the valid GPS signal is reflected multiple times off of surfaces (tree canopy, buildings) before arriving at the receiver. Since a receiver is calculating time of flight for the signal, this will add error. Good receivers have tricks to reject multipathed signals, and good antennas can also help lower their effect, but they're not magic. The other is straight up blocking of reception. GPS antennas are line of sight, and if they can't see the sky they can't get solid positions. If you're in an urban canyon, you can still see some slice of sky, and if you can lock onto 4 satellites, you can still calculate position, but you'll have a poor Horizontal Dilution of Precision. Survey grade stuff will just throw out any readings with a poor PDOP or VDOP. For like a commodity car GPS? probably take you from +/- 5 meters to +/- 30 or more, if you can even get a lock. Cell phone AGPS can help a lot in these situations.
|
# ? Feb 5, 2019 01:57 |
|
Low-Pass Filter posted:Well, the RTK base station does stay in one place and has to converge for like 20-30 minutes. Then the rover (i.e. surveyor guy with rod) can move around p much all he wants. If you're trying to locate, say, all the points on a foundation, or locate a water boundary, you're going to collect a lot of points, and you're not going to occupy them for very long. also surveyors are typically looking for elevation, not just horizontal data, which by how the system works requires more accuracy, so if you only care about your lat/long or northing/easting you can do it faster though the last time i had to go out and shoot points the station was running some kinda lovely windows xp mobile version that i spent more time wrestling with than it actually took to just calc each point
|
# ? Feb 5, 2019 02:03 |
|
ya total station software is realllll bad.
|
# ? Feb 5, 2019 02:05 |
|
Sagebrush posted:love 2 spend 20 minutes parked in the driveway spinning up the gyros and aligning my inertial platforms every time i wanna go for a drive i mean honestly if this process involved flicking rows of manual switches as indicator lights came on and you could hear stuff audibly spooling up i'd be totally down with it, that'd be rad
|
# ? Feb 5, 2019 02:59 |
|
Get a pilot's license. Even the little tiny planes are gonna have at least one or two gyroscopic instruments and you can indeed hear them spinning up and down
|
# ? Feb 5, 2019 03:48 |
|
Tesla's gonna add "Hyperdrive Autopilot" where they just play the sounds of gyros spinning and relays clicking so you think you're driving the millennium falcon
|
# ? Feb 5, 2019 04:54 |
|
Sagebrush posted:Get a pilot's license. Even the little tiny planes are gonna have at least one or two gyroscopic instruments and you can indeed hear them spinning up and down I wonder if we can get fishmech to argue a DG is a compass
|
# ? Feb 5, 2019 05:25 |
|
hobbesmaster posted:I wonder if we can get fishmech to argue yeeeeeeeeep
|
# ? Feb 5, 2019 06:35 |
|
Sagebrush posted:Get a pilot's license. lmfao no
|
# ? Feb 5, 2019 15:24 |
|
hobbesmaster posted:I wonder if we can get fishmech to argue a DG is a compass your taking us down a very dark path rn.
|
# ? Feb 5, 2019 15:40 |
|
Jabor posted:same but getting pork gyros and aligning them with my mouth year of the pig brah!
|
# ? Feb 5, 2019 17:56 |
|
|
# ? Feb 5, 2019 23:57 |
|
lol le random
|
# ? Feb 6, 2019 00:03 |
|
|
# ? Mar 29, 2024 14:12 |
|
Sagebrush posted:lol le random But that’s not a horse and it isn’t eating broccoli?
|
# ? Feb 6, 2019 00:03 |