IRL, I am a 42 year old electrical engineer. While watching your videos, I am a 12 year old kid trying to keep my jaw off the floor. You've created some really interesting projects and it is great seeing how far you've progressed over the years. Keep it up. The rap at the beginning was nerdy and wonderful too.
Being a 12 year old my jaw is totally on the floor although with inspiration from rc test flight I have since purchased a ender 3 v2 and built a cnc router and I enjoy making rc planes and drones
For efficiency science: Ask the plane to maintain an altitude that's out of ground effect and compare energy consumption against maintaining an altitude that's in ground effect? That would give you a measure like "minimum energy required to maintain level flight", which is a pretty useful metric for a ground-effect vehicle. Even if you get different speeds out of that comparison, you could still make a useful claim about energy required to travel between two waypoints.
@@gi295 which only makes it another part of the power consumption. If the plane reaches different speeds for the same "effort", then that's a base to start fine tuning the experiments. Start of with a maximum of parameters the same and measure overall consumption and speed (time between A and B), then tune parameters to achieve same speed (match the lowest of the two). Also, with such a small plane, I'm guessing it will be bouncing in and out of ground effect, which will have a slight effect on airspeed and drag too, so a variable consumption due to different drag is already happening within the same flight anyway. I agree with Saul that this would be the best approach to determine efficiency gain from ground effect.
what would be even cooler: an extremum seeking controller, maybe as an extra layer feeding a setpoint to the lower level PID ontroller, which measures energy use in real time and chooses the most efficient altitude to fly at
Would it be possible add a pressure sensor on the underside and top of the wings to measure what's going on? No idea if it would be useful, but you might be able to correlate it to efficiency, or factor it into the Ardupilot controller. Might just be a cool thing to see.
Three thoughts: 1. Awesome project, series, and music 2. Turning altitude problem: Are you rectifying the lidar altitude measurement with roll attitude? Maybe even dial in some additional compensation with roll to account for the lower wing too 3. Flying over waves: With the lidar pointing directly down, your system is reactive and not so proactive in handling disturbances in measured altitude; meaning, you measure a high wave, your controller pitches up, but you've already passed the wave. How about pointing the lidar forward at an angle to get a 'look-ahead' measurement? Better yet, mount the lidar as far forward as possible on a stick out the nose. The distance to cancel out phase delay of the wave/vehicle reaction would entirely be dependent on the length of the stick, or your total control bandwidth. Could be a fun problem to solve on video :)
Or applying a Kalman-filter like logic to the sensor signal with a override threshold for anomalies (if it's already not applied). You can do it in software with minimal computation.
I have two things that I would love to see: 1. You could create a way for the vehicle to avoid obstacles using distance sensors where it can either fly around or over things(and perhaps even calculate whether it is on a collision course. 2. A way to control the speed in ground effect by allowing the auto-pilot to adjust the pitch and throttle. Keep up the good work!!
Having a airframe that can fly efficiently both in and out of ground effect is a big deal! Being able to rise above rough water or ground obstacles and then drop back down into GF makes smaller GF vehicles much more practical.
Kavin didn't just design the Nano Goblin, but the regular Goblin, and before that the big 80 inch Graugans featured in COFPV videos. All of his planes are super efficient, and the EPP/Kevlar/thinned-Goop ones are practically indestructible.
That is a fantastic and beautiful vehicle, well done to all involved. Maybe look into creating a Kalman filter to combine all of the sensor data into a smooth-but-safe target altitude
I was thinking two sensors (nose and tail), average the readings, and filter the data to get a smoother curve and hopefully not attempt to change altitude between crests of waves.
Nice videos, but I think this would be a good idea for the next video - An RC sailboat! You can just use servos to control it and you won't need bulky batteries for the propulsion.
@@arnavpawar256 would an integrated solar panel in a rigid sail be feasible in that scenario or would the light deflection angle limit its ability to output a constant voltage?
Have you considered positioning two lidar sensors, each at the wingtip? With this you could take average out of the two, and on turns use the sensors to prevent wing strikes. My knowledge on lidar is limited so I don't for certain whether it would work.
Actually it would work, but it's not necessary. He can also make a code to increase the altitude / lidar distance value when the aircraft is turning to avoid wing strikes.
when its turning the lidar measures the hypotenuse of a right triangle so based on the turn angle and the distance measured you can calculate the distance from the ground
I think it would be better to place one sensor at the tail or in the center, and one sensor at the nose, and it should be at an angle of about 45 degrees forward. So it can tell the plane if a big wave is approaching. If it's not angled, then it would detect a big wave too late and the plane will hit the wave.
What about flared wing tips? Upwards to run parallel to the ground/water level in turn. Forcing planes higher and sort of fight the roll with increasing pressure as it rotates into the ground effect. Might make turns wider but also might add another safety against turning into the water or ground.
Just finished watching Think Flights video and was eager to see your side… and then this pops up! This has to be my favourite project yet, can’t wait to see what’s in store for us!
Those batteries amaze me, powering the plane, control surfaces, transmit video, use GPS.... amazing. Thanks for this video. I wasn't even aware that Ground Effect planes could fly so high.
Impressive! Friend of mine recently did a drone that would hold its position using an optical flow sensor and I remember how hard it was to get the control loop fine tuned. Can't even imagine how tedious that must've been for this much more dynamic system!
This is exactly the type of RC plane I could get into. ROV/Autonomy and ground effect efficiency are fascinating combinations. Wish there was a turn key or easier starter point for beginners.
Only using one LIDAR sensor is so limiting. I would use at least 2, this would allow you to average out the distance and deal with the chop, giving wrong reflection angles; you would average them and then use an offset between them to control attitude. However, I would use at least 4 sensors, adding 2 on the wingtips. This way you could easily filter out bad readings and keep from dipping your wingtips in the water. When you're at this level of expense and engineering, I would not skimp on sensors. Anyway, total hats-off to you and your collaborators and community; all of it adding up to a fantastic video experience. Thanks!
Outstanding work Daniel! On the topic of analyzing efficiency: You could integrate the power over time to get energy, and the speed to get distance. Keeping separate integrals for each mode, in-ground vs. out-of-ground effect. The ratio of total energy to total distance would then represent how far the plane could go for a given amount of energy for each mode. This should account for the nonlinearity in the relationship between speed and power. Keep the great videos coming!
This is awesome. Seeing you get the lidar on the water made me think of “dynamic soaring “. Sorta combine this project with another going for efficient flight. You always have great ideas and execution.
I think sooner or later Daniel will get approached by government agencies "explaining" to him that he can't give this knowledge to the public worldwide. This is some amazing stuff you are working on! Love this channel!
For the DJI video, i like to add 3-6db antennas for long range, It helps to have one on each side of craft so you dont lose video in turns.... Crystal patch antennas on the goggles, focus mode off, 50mb 1200mw... You should be able to go 10km with no breakup as long as ya got good antenna placement on aircraft
You are my inspiration for getting an asw28 and putting a long range fpv system on it. I have an IX 12 going to get a 10 channel smart receiver and a DJI air system for it. Hopefully it will all work together 🤞
When you were doing the ground effect in the choppy water, I was thinking how sick I would get if I was in that plane. Great shots of Seattle. I'm sure Hollywood or some Ad agency wants that footage.
I thought i recognized some of those buildings and ferries, then i looked for and saw the space needle! I didn't realize you're in western washington, maybe some time i'll catch you flying around here!
Suggestion: An autonomous plane with a radio relay in it, so you can get good video while in the ground effect at longer ranges. Perhaps it should be solar powered, or lighter-than-air so you don’t have to worry about battery as much.
I'm so proud of my home town Seattle! These videos are great and oh BTW just some awesome music production on top of it all. This is my favorite kind of entertainment straight up.
i love your projects, but what i see mostly lately ... seattle is one of the few cities in america i really want to visit. what a beautiful piece of land.
Two things come to mind. Mount the lidar module in a 2 axis swing (this way it's always pointing straight down even when banking in turns). You'd need to implement it in a way that minimizes pendulum motion when changing direction. Second thought is a bit more out there. Could you use a camera system to detect obstacles and run obstacle avoidance algorithm to navigate between boats? Maybe another LIDAR module on the top with a reflector mounted to a motor to scan 360° and build a map.
I wonder if you had two lidar modules, one in the nose and the other in the tail, and continuously averaged the two values and or used other math to smooth out the altitude signal. That way it's not trying to adjust constantly between crests and troughs of individual waves. Or additionally have some forward facing sensor to check for obstacles or approaching elevation changes in terrain. That's getting into another level of complicated, but it would be interesting if it could be more reliably autonomous.
1) adjust set point height above water in straight line flight based on wave height. calm water == lower flight. 2) put a lidar on each of the wing tips. when in a turn use the "inside" lidar as the height control input.
Takeoff is optional but landing is mandatory. The problem with WIGS is they're somewhere between the two. Survivability increases the higher you get, not lower. But terrain following might be the one thing that makes this bird fly...safely.
Really love your project! I suggest for autonomous submarine and we can see the underwater ecosystem for your next project, it would be a great journey!
Considering the efficiency test: you do not need to freeze anything, you need an integral (over time) of speed times amperage, or something like that. I do not know much about flight efficiency in particular, but it feels like an integral thing. Great video, and an awesome project. Original soundtrack is my special favorite. Edit: I wonder if it's possible to probram it to keep pitch and roll during turns in the ground effect (like, yaw-only turn) using nothing but thrust differentiation. That would exclude the wing tip touchdowns.
Your projects are really inspiring, it would be great if you could do some sort of tutorial or introduction on how to use ardupilot since there isn't a lot of information accessible.it would probably help a lot of people 🙃
How soon until Colin's amazing original music is available on Spotify? Think Flight not only designed a great airframe, but he decorated it beautifully, (not to cast any shade on your graphics, Daniel). I loved the footage of Seattle! First time I remember being there was at the World's Fair when I was a boy of 9 and I've visited many times over the years. Well done!
@rctestflight this seems as a great idea and more feasible than even terrestrial vehicles, to add to the portfolio to the companies involved in automous driven vehicles, like Tesla, Waymo, Cruise, Baidu, Mobileye, Zoox, Argo AI, Aurora, and Nuro
On some UAV's I worked on years ago, to keep the wingtips from getting in the view of the FLIR during target loiter, we used rudder in the turn and counter with alerions, to perform more of a flat turn. Slight loss of efficiency but worked well.
In the early days of LIDAR mapping we created an algorithm (like a Kalman filter) to take out tree scatter from the laser return in the heavily forested areas of the PNW. You might be able to do something similar with your wave peak and trough return to create a more average wave height calculation.
Would a LIDAR array along the X axis help provide more accurate wave/water heights, say 3 (or more) devices, front facing forward (i.e. 30 degrees), middle vertical and rear facing to the rear? Tune number of sensors, angles and separation as well as phasing. Also consider one TX/RX and two RX only?
Not only is that a great idea that came out really cool, but you even made a rap song dedicated to it! The song is just awesome 🤣🤣. Who all was involved in making the song? On a different note. You can use the goop or clear nail polish to waterproof the circuit boards. Waterproof clear nail polish is my go-to
Another great video, thanks! What I would really love is a video about the technical stuff, how you tuned the algorithms and some cool graphs and mathlab data 😀
If you can use a roll sensor to detect your roll angle accurately and make the PID ground effect control aware of the extra height required for a given roll angle using the wingspan and some trig you could bake the extra height needed for turning right into the PID control loop.
Wow… this is beautiful. The video and the music dropped my anxiety in minutes🥲 I’ve even put this in my favourites for any future needed chill sessions👌🏼😊
an extended "whisker" coming out of the bottom that could have some rigidity, with an actual water sensor could sample under the wave tips for contiguos signal
I'm curious if a forward positioned lidar could be used to help anticipate upcoming ground / water changes in elevation? Regardless, this is awesome. Great work!
Super cool Daniel!... Thanks to your inspiration I am working a fpv rc plane, where the controls and video feed is handled via raspberry pi's and 4g networks... Obviously one wouldn't be able to fly the thing if there's no cellular connection, but I personally think it's a good challenge. Plus where I am from Fpv stuff isn't cheap as far as I can tell
IRL, I am a 42 year old electrical engineer. While watching your videos, I am a 12 year old kid trying to keep my jaw off the floor. You've created some really interesting projects and it is great seeing how far you've progressed over the years. Keep it up. The rap at the beginning was nerdy and wonderful too.
Great Video
Play some dead or alive to the video
I'm a 51 year old CS & EE. Makes me wish I had been so innovative in those earlier years. This guy is living a dream life, doing what you love.
Being a 12 year old my jaw is totally on the floor although with inspiration from rc test flight I have since purchased a ender 3 v2 and built a cnc router and I enjoy making rc planes and drones
You’re the guy from split
"High-Quality Autonomous Ground Effect Vehicle" has been added to my spotify playlist
Fck that song haha
i need it
Best music
[Insert guy looking at smoking airpod here]
It's unreal
another banger of a track, ground effect the album when?
Sick track for sure. One request. Add heavy bass, thats worth some public disturbance 🔊 🔊🔉🔉🔉🔊
Collaboration.?
never thought the word "ardupilot" would be used in a song lmao. amazing
@John Thomas Bro jealousy all over you 😂. Why even take the time to comment? If you don't like it gtfo
@John Thomas Nobody cares
Everything about this project is soooooo impressive!
You two guys are absolute legends!
Yeah, this plane is Awesome!
the song most of all
yes
WE HAVE YET KNOWLEDGE FLAT EARTH
For efficiency science: Ask the plane to maintain an altitude that's out of ground effect and compare energy consumption against maintaining an altitude that's in ground effect? That would give you a measure like "minimum energy required to maintain level flight", which is a pretty useful metric for a ground-effect vehicle. Even if you get different speeds out of that comparison, you could still make a useful claim about energy required to travel between two waypoints.
The faster you go, the more the drag. The difference in efficiency because of speed is very substantial, more so than the ground effect
@@gi295 which only makes it another part of the power consumption. If the plane reaches different speeds for the same "effort", then that's a base to start fine tuning the experiments. Start of with a maximum of parameters the same and measure overall consumption and speed (time between A and B), then tune parameters to achieve same speed (match the lowest of the two). Also, with such a small plane, I'm guessing it will be bouncing in and out of ground effect, which will have a slight effect on airspeed and drag too, so a variable consumption due to different drag is already happening within the same flight anyway.
I agree with Saul that this would be the best approach to determine efficiency gain from ground effect.
what would be even cooler: an extremum seeking controller, maybe as an extra layer feeding a setpoint to the lower level PID ontroller, which measures energy use in real time and chooses the most efficient altitude to fly at
Would it be possible add a pressure sensor on the underside and top of the wings to measure what's going on? No idea if it would be useful, but you might be able to correlate it to efficiency, or factor it into the Ardupilot controller. Might just be a cool thing to see.
Actually it won't be that useful as it would be hard to set it to a value. But that's a nice suggestion.
I think it would be cool to see when the pressure semi stabilizes and ground effect is achieved
@@ApacheFPV yep
That's how an angle of attack indicator gets it's signals. It has to be calibrated one time in flight at zero g, if I remember how I calibrated my RV7
It is fine if the ground is flat. But with waves it is pretty hard to determinate what is the best pressure in this small scale.
These autonomous vehicle videos are fantastic, keep it up!
ruclips.net/video/jXP6YwH6o1ki/видео.htmluy
Boring one trick pony
these songs slap harder and harder every time
lol true
Honestly prefer just the instrumental. End up muting it every time or skipping the time lapses
Three thoughts:
1. Awesome project, series, and music
2. Turning altitude problem: Are you rectifying the lidar altitude measurement with roll attitude? Maybe even dial in some additional compensation with roll to account for the lower wing too
3. Flying over waves: With the lidar pointing directly down, your system is reactive and not so proactive in handling disturbances in measured altitude; meaning, you measure a high wave, your controller pitches up, but you've already passed the wave. How about pointing the lidar forward at an angle to get a 'look-ahead' measurement? Better yet, mount the lidar as far forward as possible on a stick out the nose. The distance to cancel out phase delay of the wave/vehicle reaction would entirely be dependent on the length of the stick, or your total control bandwidth. Could be a fun problem to solve on video :)
I love your work, I might try your flight controller and autopilot code some day if I have a chance.
I was wondering if it would be possible to mount it to a gimbal of sorts. Waterproofing would be an issue though.
Or several LIDAR units across the wingspan? Average them out with bias towards the central one in order to reduce input noise.
Or applying a Kalman-filter like logic to the sensor signal with a override threshold for anomalies (if it's already not applied). You can do it in software with minimal computation.
I have two things that I would love to see: 1. You could create a way for the vehicle to avoid obstacles using distance sensors where it can either fly around or over things(and perhaps even calculate whether it is on a collision course.
2. A way to control the speed in ground effect by allowing the auto-pilot to adjust the pitch and throttle.
Keep up the good work!!
ruclips.net/video/jXP6YwH6o1ku/видео.htmlt4
Having a airframe that can fly efficiently both in and out of ground effect is a big deal! Being able to rise above rough water or ground obstacles and then drop back down into GF makes smaller GF vehicles much more practical.
Kavin didn't just design the Nano Goblin, but the regular Goblin, and before that the big 80 inch Graugans featured in COFPV videos. All of his planes are super efficient, and the EPP/Kevlar/thinned-Goop ones are practically indestructible.
That is a fantastic and beautiful vehicle, well done to all involved.
Maybe look into creating a Kalman filter to combine all of the sensor data into a smooth-but-safe target altitude
ruclips.net/video/jXP6YwH6o1ku/видео.htmlt5
I was thinking two sensors (nose and tail), average the readings, and filter the data to get a smoother curve and hopefully not attempt to change altitude between crests of waves.
I forgot about this channel. Come back every couple years. Every time you’re so much further than you were. So impressive!
Nice videos, but I think this would be a good idea for the next video - An RC sailboat!
You can just use servos to control it and you won't need bulky batteries for the propulsion.
yes.
something like "saildrone"
As it would need very small batteries for the sailboat, he can use solar panels for the recharging so it would have infinite range.
@@arnavpawar256 would an integrated solar panel in a rigid sail be feasible in that scenario or would the light deflection angle limit its ability to output a constant voltage?
@@AirJimInCT you could prabably have flexible solar panels on the sail, which might charge with enough heel due to the wind?
Have you considered positioning two lidar sensors, each at the wingtip? With this you could take average out of the two, and on turns use the sensors to prevent wing strikes. My knowledge on lidar is limited so I don't for certain whether it would work.
Actually it would work, but it's not necessary. He can also make a code to increase the altitude / lidar distance value when the aircraft is turning to avoid wing strikes.
when its turning the lidar measures the hypotenuse of a right triangle so based on the turn angle and the distance measured you can calculate the distance from the ground
I think it would be better to place one sensor at the tail or in the center, and one sensor at the nose, and it should be at an angle of about 45 degrees forward. So it can tell the plane if a big wave is approaching. If it's not angled, then it would detect a big wave too late and the plane will hit the wave.
What about flared wing tips? Upwards to run parallel to the ground/water level in turn. Forcing planes higher and sort of fight the roll with increasing pressure as it rotates into the ground effect. Might make turns wider but also might add another safety against turning into the water or ground.
When on ground effect, can turning with just rudder adjustment an option?
Just finished watching Think Flights video and was eager to see your side… and then this pops up! This has to be my favourite project yet, can’t wait to see what’s in store for us!
ruclips.net/video/jXP6YwH6o1ky/видео.htmlt3
Solar plane V3!
Those batteries amaze me, powering the plane, control surfaces, transmit video, use GPS.... amazing. Thanks for this video. I wasn't even aware that Ground Effect planes could fly so high.
You're audio musical interludes are total cheese and absolutely based. Keep that shit up!!
Impressive! Friend of mine recently did a drone that would hold its position using an optical flow sensor and I remember how hard it was to get the control loop fine tuned. Can't even imagine how tedious that must've been for this much more dynamic system!
The song is solid contender for the European Song Contest 2022!
This is exactly the type of RC plane I could get into. ROV/Autonomy and ground effect efficiency are fascinating combinations. Wish there was a turn key or easier starter point for beginners.
I come for the autonomous vehicles and stay for the songs!
Only using one LIDAR sensor is so limiting. I would use at least 2, this would allow you to average out the distance and deal with the chop, giving wrong reflection angles; you would average them and then use an offset between them to control attitude.
However, I would use at least 4 sensors, adding 2 on the wingtips. This way you could easily filter out bad readings and keep from dipping your wingtips in the water. When you're at this level of expense and engineering, I would not skimp on sensors.
Anyway, total hats-off to you and your collaborators and community; all of it adding up to a fantastic video experience. Thanks!
Just another heads up - your style of editing and using footage is magic, my friend. I really enjoy when you post.
one of the best parts of these videos is seeing all those random people you meet like the paddle boarder guy or the guy that took you on his boat
Love how you guys went from an Alexeyev style Ekranoplan to a Bartini styled GEV as it gets better and better. Shows the design priority changing.
ruclips.net/video/jXP6YwH6o1ku/видео.htmly5
The fact that the breadcrumb trail is an A-6 intruder silhouette is a nice touch
Outstanding work Daniel! On the topic of analyzing efficiency: You could integrate the power over time to get energy, and the speed to get distance. Keeping separate integrals for each mode, in-ground vs. out-of-ground effect. The ratio of total energy to total distance would then represent how far the plane could go for a given amount of energy for each mode. This should account for the nonlinearity in the relationship between speed and power. Keep the great videos coming!
I'm a huge GEV fan since the soviet ekranoplans and I actually had to watch this in parts not to be overwhelmed by excitement. Wow.
I watch this channel for both the tech and the scenery in equal amounts.
This is awesome. Seeing you get the lidar on the water made me think of “dynamic soaring “. Sorta combine this project with another going for efficient flight.
You always have great ideas and execution.
This channel has been consistently putting out some of the best projects I have seen for years.
ruclips.net/video/jXP6YwH6o1ku/видео.htmlt3
@@Hitycooking Moving up and out of the third world with the internet. Respect the hustle.
Wow, what a beautiful background you have there with that snow-covered mountain!
i don’t know what’s more epic. the plane or the self written soundtrack 👍🏼
Thanks for blowing the spider off, and not killing it!
I have been watching your videos for years, and I still can find nothing more relaxing than this.
I think sooner or later Daniel will get approached by government agencies "explaining" to him that he can't give this knowledge to the public worldwide. This is some amazing stuff you are working on! Love this channel!
I love the take-off shots, showing the rippling water.
The self made music is so bad, I love it, dont change it. It is fun, actually puts a smile on my face
For the DJI video, i like to add 3-6db antennas for long range, It helps to have one on each side of craft so you dont lose video in turns.... Crystal patch antennas on the goggles, focus mode off, 50mb 1200mw... You should be able to go 10km with no breakup as long as ya got good antenna placement on aircraft
16:20, loved blocking up and playing around with that creek just behind you, probably still just as fun
Not sure how this video showed up in my feed but glad it. I live in seattle area also. Would be awesome to see these guys doing this in real life.
This is my second rctestflight video, I respect the amount of time you put into your videos.
You are my inspiration for getting an asw28 and putting a long range fpv system on it.
I have an IX 12 going to get a 10 channel smart receiver and a DJI air system for it.
Hopefully it will all work together 🤞
When you were doing the ground effect in the choppy water, I was thinking how sick I would get if I was in that plane.
Great shots of Seattle. I'm sure Hollywood or some Ad agency wants that footage.
That shot of Seattle around 18 minutes was amazing
Man, I just discovered TF and binged his vids last week. Now the rctestflight vid is out. Sweet.
Those time lapses at the beginning of the video had me straight anxious about a collision
I'm a simple person - I see rctestflight - I like and share
I thought i recognized some of those buildings and ferries, then i looked for and saw the space needle! I didn't realize you're in western washington, maybe some time i'll catch you flying around here!
You really really outdid yourself. Bravo! And I love the music and lyrics!
This is my new favorite RUclips channel. Keep being awesome dude. I love what you're doing.
Dude! Music is NEXT LEVEL! In all seriousness, your content is top notch! Appreciate you and your crew!
Super cool, it needs a way to sense for obstacles and make adjustments to it's preprogrammed course.
Suggestion: An autonomous plane with a radio relay in it, so you can get good video while in the ground effect at longer ranges.
Perhaps it should be solar powered, or lighter-than-air so you don’t have to worry about battery as much.
Future airliner designer right here, ladies and gentlemen!!
I'm so proud of my home town Seattle! These videos are great and oh BTW just some awesome music production on top of it all. This is my favorite kind of entertainment straight up.
hahaha I've been looking forward to another banger this hot new artist. Good work boys!
I really like your musical segments when they are short.
As others have said, this is an awesome project. And the "ground effect" music.. A super nice touch and I love it!
These videos are always fun to watch because I live in seattle.
i love your projects, but what i see mostly lately ... seattle is one of the few cities in america i really want to visit. what a beautiful piece of land.
Two things come to mind. Mount the lidar module in a 2 axis swing (this way it's always pointing straight down even when banking in turns). You'd need to implement it in a way that minimizes pendulum motion when changing direction.
Second thought is a bit more out there. Could you use a camera system to detect obstacles and run obstacle avoidance algorithm to navigate between boats? Maybe another LIDAR module on the top with a reflector mounted to a motor to scan 360° and build a map.
Everyone that likes this channel will like Think Flight. Great planes, experiments and videos.
Your music drops are the best. …when you’re flying only one inch off the ground…
I'm always amazed seeing the hardwork you do on that videos :) keep donig it, it's awesome
I wonder if you had two lidar modules, one in the nose and the other in the tail, and continuously averaged the two values and or used other math to smooth out the altitude signal. That way it's not trying to adjust constantly between crests and troughs of individual waves. Or additionally have some forward facing sensor to check for obstacles or approaching elevation changes in terrain. That's getting into another level of complicated, but it would be interesting if it could be more reliably autonomous.
The music alone deserves a like. About the vehicle: impressive
1) adjust set point height above water in straight line flight based on wave height. calm water == lower flight.
2) put a lidar on each of the wing tips. when in a turn use the "inside" lidar as the height control input.
Came for the project, stayed for the music!
Takeoff is optional but landing is mandatory. The problem with WIGS is they're somewhere between the two. Survivability increases the higher you get, not lower. But terrain following might be the one thing that makes this bird fly...safely.
Really love your project! I suggest for autonomous submarine and we can see the underwater ecosystem for your next project, it would be a great journey!
As a Seattlite also into flying machines this is so amazing. I hope to meet you one day!
Considering the efficiency test: you do not need to freeze anything, you need an integral (over time) of speed times amperage, or something like that. I do not know much about flight efficiency in particular, but it feels like an integral thing.
Great video, and an awesome project. Original soundtrack is my special favorite.
Edit: I wonder if it's possible to probram it to keep pitch and roll during turns in the ground effect (like, yaw-only turn) using nothing but thrust differentiation. That would exclude the wing tip touchdowns.
Your projects are really inspiring, it would be great if you could do some sort of tutorial or introduction on how to use ardupilot since there isn't a lot of information accessible.it would probably help a lot of people 🙃
How soon until Colin's amazing original music is available on Spotify? Think Flight not only designed a great airframe, but he decorated it beautifully, (not to cast any shade on your graphics, Daniel). I loved the footage of Seattle! First time I remember being there was at the World's Fair when I was a boy of 9 and I've visited many times over the years. Well done!
@rctestflight this seems as a great idea and more feasible than even terrestrial vehicles, to add to the portfolio to the companies involved in automous driven vehicles, like Tesla, Waymo, Cruise, Baidu, Mobileye, Zoox, Argo AI, Aurora, and Nuro
There's no way you can't like the vid once you hear those bars 🔥
On some UAV's I worked on years ago, to keep the wingtips from getting in the view of the FLIR during target loiter, we used rudder in the turn and counter with alerions, to perform more of a flat turn. Slight loss of efficiency but worked well.
In the early days of LIDAR mapping we created an algorithm (like a Kalman filter) to take out tree scatter from the laser return in the heavily forested areas of the PNW. You might be able to do something similar with your wave peak and trough return to create a more average wave height calculation.
Would a LIDAR array along the X axis help provide more accurate wave/water heights, say 3 (or more) devices, front facing forward (i.e. 30 degrees), middle vertical and rear facing to the rear? Tune number of sensors, angles and separation as well as phasing. Also consider one TX/RX and two RX only?
Not only is that a great idea that came out really cool, but you even made a rap song dedicated to it! The song is just awesome 🤣🤣. Who all was involved in making the song?
On a different note. You can use the goop or clear nail polish to waterproof the circuit boards. Waterproof clear nail polish is my go-to
So rad! I hope you can find some nice snowy fields to try it over this winter!
Came here for the plane, stayed for the sick beats 🔥
Another great video, thanks! What I would really love is a video about the technical stuff, how you tuned the algorithms and some cool graphs and mathlab data 😀
You or the person that does your songs should make an album for us!
When does the album drop?
If you can use a roll sensor to detect your roll angle accurately and make the PID ground effect control aware of the extra height required for a given roll angle using the wingspan and some trig you could bake the extra height needed for turning right into the PID control loop.
Now you need radar to detect oncoming objects and this will be unstoppable! Awesome work.
Great technical video as usual...The lake frozen in Winter is going to be interesting, especially combined with cold, calm dead air...🤔😀😀🇬🇧🇬🇧
That song is a banger and a good narrative device.
Super cool plane and video, loved the collaboration and following the Think Flight channel too
Wow… this is beautiful. The video and the music dropped my anxiety in minutes🥲 I’ve even put this in my favourites for any future needed chill sessions👌🏼😊
an extended "whisker" coming out of the bottom that could have some rigidity, with an actual water sensor could sample under the wave tips for contiguos signal
I'm curious if a forward positioned lidar could be used to help anticipate upcoming ground / water changes in elevation? Regardless, this is awesome. Great work!
that plane flies as smooth as BUTTER! Love it!
best footage so far! legitimately amazing shots in the later half of the video
Lol I could see my apartment from one of your aerial shots. Beautiful cinematography as well as great science.
Greetings from Spokane. Absolutely Sweet!
Boy, I wish that I lived a LOT closer. No skill (just old) but handy and interested to learning.
Super cool Daniel!... Thanks to your inspiration I am working a fpv rc plane, where the controls and video feed is handled via raspberry pi's and 4g networks...
Obviously one wouldn't be able to fly the thing if there's no cellular connection, but I personally think it's a good challenge. Plus where I am from Fpv stuff isn't cheap as far as I can tell