Open frednickler opened 4 years ago
baykah
commented
4 years ago Not really :(, last thing I did was to use a 10Hz GPS on the TAG, but in the end I kind lost intrest.
My coding skill back then was worst than today, I think some thing needs to be rewriten, and I never succeded in applying a 2D kallman filter (position + velocity) which I beleive is (plus or minus) one key element.
I'm prety sure "solo shot" (search on google) use laso pressure for altitude.
frednickler
commented
4 years ago Your idea with the Kalman filter was to improve the accuracy or to use it in case the base lost connection with the transmitter? Also, to make that work you would have to add a separate "MPU-6050 Module 3 Axis Gyroscope + Accelerometer" to the base, correct?
I read you wrote your own serial comm protocol for the 2 boards, would you be able to share that please?
Regarding Soloshot, I know the product but don't really know how the software and hardware really work. I have looked online, but there is not much detailed info. Having said that I found this link and it contains some useful information. So I will post below in case some comes looking for it.
ahmetcemturan
commented
3 years ago Hi @baykah,
Very interested in your project. I am really new to Arduino but could provide help for CAD.
Would You be interested to develop it further?
A few thoughts:
-Altitude : I am not sure soloshot uses any pressure sensor (not sure what You mean with "laso pressure") as their beacons need to be sealed as they are used on-in the water too. I myself want it for kitesurfing. When You look at footage for RC Airplane mode You see that it is succesful trying to follow in X/Y but not in Z. The thing that You have to choose explicitly RC-Airplane mode makes me think that it is having a AI thingy built in that tracks "the object in the sky that is in that general X/Y direction". Other than that You solution is much more successful in tracking than theirs. I believe that a workaround around hermetically sealing could be, to put the pressure sensor behind a membrane. Then at initialisation both beacon and receiver could compare and set their baseline so that any difference in internal pressure is egalised.
-Jerky movement: I believe printed harmonic gears could work well for the movement. They are backlash free and not backdriveable. I have experimented with TPU flexsplines and found they work really well. Reductions up to 1:50 would be possible that way.
-Offset for CMOS location: I haven't put any thought how it would work but I think that there should be some way/formulation to offset the CMOS from the X/Y/Z pivot point in software rather than to have to re design mounting for each different camera/setup.
Would really be interested in pushing this project forward. I think I could also get some Programming help.
baykah
commented
3 years ago Hello,
I stopped working on this project a few years ago now ...
I think the main issue is software, the jerky movement comes from there 😕 It's lacking a proper kallman 2D filter (pos + speed), to be able to drive the servo in a faster loop (I think 😄)
Last thing I did was to acquire a 10Hz GPS, but it had no internal memory and got plenty of problem of configuration reseting, then I stopped.
I would need to get back to the project by buying a better 10Hz GPS and do a complete software rewrite.
Regarding altitude, if it is not using pressure but only GPS, it a matter of GPS performance + complexe software. I switch to pressure when I wanted accuracy, I thought that my Z problem during film attemps was link also to software...
Regards
De : Ahmet Cem TURAN @.> Envoyé : vendredi 9 avril 2021 08:52 À : baykah/Autonomous-video-recording @.> Cc : baykah @.>; Mention @.> Objet : Re: [baykah/Autonomous-video-recording] Updates? (#1)
Hi @baykahhttps://github.com/baykah, Very interested in your project. I am really new to Arduino but could provide help for CAD. Would You be interested to develop it further? A few thoughts: -Altitude : I am not sure soloshot uses any pressure sensor (not sure what You mean with "laso pressure") as their beacons need to be sealed as they are used on-in the water too. I myself want it for kitesurfing. When You look at footage for RC Airplane mode You see that it is succesful trying to follow in X/Y but not in Z. The thing that You have to choose explicitly RC-Airplane mode makes me think that it is having a AI thingy built in that tracks "the object in the sky that is in that general X/Y direction". Other than that You solution is much more successful in tracking than theirs. I believe that a workaround around hermetically sealing could be, to put the pressure sensor behind a membrane. Then at initialisation both beacon and receiver could compare and set their baseline so that any difference in internal pressure is egalised. -Jerky movement: I believe printed harmonic gears could work well for the movement. They are backlash free and not backdriveable. I have experimented with TPU flexsplines and found they work really well. Reductions up to 1:50 would be possible that way. -Offset for CMOS location: I haven't put any thought how it would work but I think that there should be some way/formulation to offset the CMOS from the X/Y/Z pivot point in software rather than to have to re design mounting for each different camera/setup.
Would really be interested in pushing this project forward. I think I could also get some Programming help.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/baykah/Autonomous-video-recording/issues/1#issuecomment-816456204, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ADC3UTTEM2MMAOVBESPDST3TH2P3RANCNFSM4MSQLAAA.
baykah
commented
3 years ago I bought a MTK3339 on aliexpress to see if I want to resurect this project ...
Idealy if it's working well, I succed in rewriting the code and implement good kallman 2D filter I will buy a second GPS and try to get it to work again.
Maybe also try to get it work with the altitude with the GPS ..
De : Ahmet Cem TURAN @.> Envoyé : vendredi 9 avril 2021 08:52 À : baykah/Autonomous-video-recording @.> Cc : baykah @.>; Mention @.> Objet : Re: [baykah/Autonomous-video-recording] Updates? (#1)
Hi @baykahhttps://github.com/baykah, Very interested in your project. I am really new to Arduino but could provide help for CAD. Would You be interested to develop it further? A few thoughts: -Altitude : I am not sure soloshot uses any pressure sensor (not sure what You mean with "laso pressure") as their beacons need to be sealed as they are used on-in the water too. I myself want it for kitesurfing. When You look at footage for RC Airplane mode You see that it is succesful trying to follow in X/Y but not in Z. The thing that You have to choose explicitly RC-Airplane mode makes me think that it is having a AI thingy built in that tracks "the object in the sky that is in that general X/Y direction". Other than that You solution is much more successful in tracking than theirs. I believe that a workaround around hermetically sealing could be, to put the pressure sensor behind a membrane. Then at initialisation both beacon and receiver could compare and set their baseline so that any difference in internal pressure is egalised. -Jerky movement: I believe printed harmonic gears could work well for the movement. They are backlash free and not backdriveable. I have experimented with TPU flexsplines and found they work really well. Reductions up to 1:50 would be possible that way. -Offset for CMOS location: I haven't put any thought how it would work but I think that there should be some way/formulation to offset the CMOS from the X/Y/Z pivot point in software rather than to have to re design mounting for each different camera/setup.
Would really be interested in pushing this project forward. I think I could also get some Programming help.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/baykah/Autonomous-video-recording/issues/1#issuecomment-816456204, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ADC3UTTEM2MMAOVBESPDST3TH2P3RANCNFSM4MSQLAAA.
ahmetcemturan
commented
3 years ago I am willing to provide any kind of help You need. (Within my capabilities.)12:50, April 9, 2021, baykah @.***>: I bought a MTK3339 on aliexpress to see if I want to resurect this project ...
Idealy if it's working well, I succed in rewriting the code and implement good kallman 2D filter I will buy a second GPS and try to get it to work again.
Maybe also try to get it work with the altitude with the GPS ..
De : Ahmet Cem TURAN @.> Envoyé : vendredi 9 avril 2021 08:52 À : baykah/Autonomous-video-recording @.> Cc : baykah @.>; Mention @.> Objet : Re: [baykah/Autonomous-video-recording] Updates? (#1)
Hi @baykahhttps://github.com/baykah, Very interested in your project. I am really new to Arduino but could provide help for CAD. Would You be interested to develop it further? A few thoughts: -Altitude : I am not sure soloshot uses any pressure sensor (not sure what You mean with "laso pressure") as their beacons need to be sealed as they are used on-in the water too. I myself want it for kitesurfing. When You look at footage for RC Airplane mode You see that it is succesful trying to follow in X/Y but not in Z. The thing that You have to choose explicitly RC-Airplane mode makes me think that it is having a AI thingy built in that tracks "the object in the sky that is in that general X/Y direction". Other than that You solution is much more successful in tracking than theirs. I believe that a workaround around hermetically sealing could be, to put the pressure sensor behind a membrane. Then at initialisation both beacon and receiver could compare and set their baseline so that any difference in internal pressure is egalised. -Jerky movement: I believe printed harmonic gears could work well for the movement. They are backlash free and not backdriveable. I have experimented with TPU flexsplines and found they work really well. Reductions up to 1:50 would be possible that way. -Offset for CMOS location: I haven't put any thought how it would work but I think that there should be some way/formulation to offset the CMOS from the X/Y/Z pivot point in software rather than to have to re design mounting for each different camera/setup.
Would really be interested in pushing this project forward. I think I could also get some Programming help.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/baykah/Autonomous-video-recording/issues/1#issuecomment-816456204, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ADC3UTTEM2MMAOVBESPDST3TH2P3RANCNFSM4MSQLAAA.
—You are receiving this because you commented.Reply to this email directly, view it on GitHub, or unsubscribe. -- Sent from Yandex.Mail for mobile
baykah
commented
3 years ago I'm pretty sure everything is on github if you want to work on your end.
I'm pretty sure soloshot also use gyro+accelerometer to provide more information of the beacon than the simple GPS ....
Basicly, soloshot must have years of software developpement to get it work, the rest is "simple" mechanics.
I may work on this when I get the new GPS, but what's lacking is a talented software engineer ...
De : Ahmet Cem TURAN @.> Envoyé : vendredi 9 avril 2021 19:37 À : baykah/Autonomous-video-recording @.> Cc : baykah @.>; Mention @.> Objet : Re: [baykah/Autonomous-video-recording] Updates? (#1)
I am willing to provide any kind of help You need. (Within my capabilities.)12:50, April 9, 2021, baykah @.***>: I bought a MTK3339 on aliexpress to see if I want to resurect this project ...
Idealy if it's working well, I succed in rewriting the code and implement good kallman 2D filter I will buy a second GPS and try to get it to work again.
Maybe also try to get it work with the altitude with the GPS ..
De : Ahmet Cem TURAN @.> Envoyé : vendredi 9 avril 2021 08:52 À : baykah/Autonomous-video-recording @.> Cc : baykah @.>; Mention @.> Objet : Re: [baykah/Autonomous-video-recording] Updates? (#1)
Hi @baykahhttps://github.com/baykah, Very interested in your project. I am really new to Arduino but could provide help for CAD. Would You be interested to develop it further? A few thoughts: -Altitude : I am not sure soloshot uses any pressure sensor (not sure what You mean with "laso pressure") as their beacons need to be sealed as they are used on-in the water too. I myself want it for kitesurfing. When You look at footage for RC Airplane mode You see that it is succesful trying to follow in X/Y but not in Z. The thing that You have to choose explicitly RC-Airplane mode makes me think that it is having a AI thingy built in that tracks "the object in the sky that is in that general X/Y direction". Other than that You solution is much more successful in tracking than theirs. I believe that a workaround around hermetically sealing could be, to put the pressure sensor behind a membrane. Then at initialisation both beacon and receiver could compare and set their baseline so that any difference in internal pressure is egalised. -Jerky movement: I believe printed harmonic gears could work well for the movement. They are backlash free and not backdriveable. I have experimented with TPU flexsplines and found they work really well. Reductions up to 1:50 would be possible that way. -Offset for CMOS location: I haven't put any thought how it would work but I think that there should be some way/formulation to offset the CMOS from the X/Y/Z pivot point in software rather than to have to re design mounting for each different camera/setup.
Would really be interested in pushing this project forward. I think I could also get some Programming help.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/baykah/Autonomous-video-recording/issues/1#issuecomment-816456204, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ADC3UTTEM2MMAOVBESPDST3TH2P3RANCNFSM4MSQLAAA.
—You are receiving this because you commented.Reply to this email directly, view it on GitHub, or unsubscribe. -- Sent from Yandex.Mail for mobile
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/baykah/Autonomous-video-recording/issues/1#issuecomment-816843475, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ADC3UTWBFDIYNUTI7MPYSK3TH43OBANCNFSM4MSQLAAA.
ahmetcemturan
commented
3 years ago For servo easing I found this:https://youtu.be/fC9uxdOBhfA
Hi @baykah
Are you still working on the this project? Have you finished or updated it? I would like to start a similar project and would use your project as guidance. Thanks.