3D Printed Drone Boat With PixHawk Flavored Sprinkles On Top
ฝัง
- เผยแพร่เมื่อ 17 พ.ค. 2024
- Sponsored by PCBWay.
When I was young, me and my dad used to go out sailing together. This project channels my childhood into a classic aka: Matchstic project - an autonomous drone boat. Its 3D printed, using my usual PixHawk and Raspberry Pi combo for control.
Its prototype number 1 of my attempt at the microtransat, an international competion to send an autonomous or unmanned vessel across the Atlantic Ocean. Gotta set those ambitions high, huh!
🌐 Links
PCBWay - Video sponsor
www.pcbway.com
Microtransat
www.microtransat.org/
How to connect Pixhawk and Raspberry Pi (telemetry cable, and using MAVProxy)
www.hackster.io/Matchstic/con...
Tuning ArduPilot (rover docs, but the same for boats)
ardupilot.org/rover/docs/rove...
3D designs for the hull (missing velcro strap blocks)
www.dropbox.com/scl/fo/6aqov1...
😎 Social Stuff
Mastodon: mastodon.online/@akamatchstic
Patreon: / akamatchstic
🎞️ Chapters
00:00 - Introduction
01:02 - Building the hull
05:48 - Electronics
06:58 - Bathtub test
08:57 - Lake tests
12:03 - Tuning
14:22 - Final test
ℹ️ Attributions
Samuel Fielder - Woodstock's second sea trip at Weston
• Woostock's second sea ...
🎵 Music
In order of usage:
bezserca - Purpose
• bezserca - purpose
TrackTribe - Upstate
• TrackTribe - "Upstate"...
DivKid - Icelandic Arpeggios
• Icelandic Arpeggios
Punch Deck - Neon Underworld
• Punch Deck - Neon Unde...
Alex-Productions - Startup
• Minimal Technology Cor...
Punch Deck - Persistence
• Punch Deck - Persistence
BVG x møndberg - insomnia
• BVG x møndberg - insomnia
nihilore - Garden
www.nihilore.com/prog-rock
Depth Fuse - French Fuse
• Depth Fuse
Punch Deck - Elegance in Simplicity
• Punch Deck - Elegance ...
Keywords: ArduPilot, PixHawk, Pix32, microtransat - วิทยาศาสตร์และเทคโนโลยี
The Atlantic ocean is about 10.000 times more unforgiving than you seem to aim for. Please build your next version like a tank, then double all your specs, then double them again. I have a long-term project to have an autonomous vehicle cross the pacific ocean from Tokyo to Seattle (not public yet), I'll be following your journey. Keep up ! 👊
I love the way you journal. The audience is doing this with you.
Looking great, can tell you've put a lot of work in.
An Atlantic voyage will be quite a change from quiet lakes but then it has to start somewhere - can't wait to see how you get on!
Great stuff :)
Ooh I'm excited to see this in the open sea!
You and me both! Got a couple more (working) prototypes to go before I reckon 😅
found your channel by accident, now watching fourth video in a row. Keep up with a good quality content!
Welcome to the channel! 😅
Finally 🎉
The version for open sea will definetly needs to be more robust and made from single piece and already hard prooven proffesional (seal) parts. I do not trust ESP32, there needs to be also some controller redundancy. Maybe if this will be bigget boat with two 200Wp photovoltaic panels, it should be dueble.
Definitely agreed! Looking to move to a mono hull design built from a fiberglass layup, using a radio link for transmitter and satellite internet for eg GPS updates when out at sea. Got a lot of learning to go before then 😅
bird,shark,wave proof?
Congratulations on the sensational content! My students (6th Elementary School) watched and are very inspired to try to build a drone boat (obviously, much simpler), but inspired by their work. So please don't give up: your audience awaits your return!
I’d say they should go for it! A drone boat can be as simple as a plastic box with some holes where the propellers come out 😊
Aiming to continue when I can! Got a huge amount of renovation work coming up with a house move, and had the workshop packed for months - lots of stress and delays, but finally looking like it’s happening soon
How will your boat right itself if it is flipping upsidedown by a wave. BTW, you would do well to have a bilge pump and seal your electronics with a marine grade epoxy. They don't get along well with salt water. If a 3D printed hull is your final design, then cover the entire exterior with fiberglass cloth and epoxy to prevent a hull breach and achieve as low a drag as possible. Lastly you need to make major improvements to the area of the propellor shaft seals and glue the u-joints to prevent loosening while underway. Good luck.
Wow that was good! I know that you can do it Matt! I belive in you!
You've given yourself a huge challenge, but I am sure you will succeed. Looking forward to future vids.
Cool Colin Furze shirt, you have good taste. Great job!
I absolutely love your content! I came across your videos while browsing drone footage, and I must say, it's some really cool stuff you've got going on.
I've been trying to find your email but couldn't locate it. I wanted to ask for your advice on how to go about learning all the skills needed to fly an autonomous drone. You see, I'm an electronics engineer, and I'm very keen on making my graduation project an autonomous drone, although I'm still unsure about its exact purpose.
Could you kindly provide some insight on where to begin? Perhaps suggest some courses, books, or resources to get started? Your help would mean the world to me. Thanks a ton for the incredible content you create!
Awesome work Matt
Are you going to pull in data from vessel finder to help avoid a collision in the shipping lanes?
Like your videos and instructions. Great job!
Afraid your boat will fail tho, props will clog. This ain't an easy challenge.
This current prototype wouldn’t make it for sure - got a lot of work to go before an attempt
Nice work. Do the motors spin in the same direction?
If it's not underway yet, (1) look at the newest development in propellers where props don't have solid blades but loops. These can be much more efficient. You'll be aware that currents in the Atlantic can easily flow at a couple knots and when these are against your boat, you'll need to be able to exceed that in order to get forward. (1.1) this means your waypoints nav AI may need to be able to adapt its route to those current patterns. Along such lines, if you are familiar with AIS and sailors on the fear of ocean going sailboats of being overrun by large cargo/cruise (military?) ships then you'll know how they use AIS to sleep better and occasionally steer clear of a collision. Some of these sailors have called a, say, 400m long container freight ship to ascertain their actual path - they got politely answered but the crew in the wheelhouse may have been ROTFL after that call.
Then, what if (2) heavier ocean waves flip the boat over? Is it self-righting? If the boat doesn't have sensors that measure its orientation (and wave motion from that) and so it cannot use such data to sail perpendicular to the rolling direction of waves, then trying to prevent the flips is difficult and getting flipped over is a very serious risk.
The boat can navigate waypoints as you demonstrate but when it is upside down in the water, collects no solar, and tries to navigate the waypoints, the battery will be flat really quickly. So when another wave flips it in the right position again, will it (3) boot its electronics again once the solar has collected enough power?
The "big" lid is (4) a bit of a leak risk as you demonstrate and I probably would have opted for a (5) two compartment design where the motors are in a separate chamber, so there would be a waterwall (cf. firewall) between the electronics and the motors. This would also make the lids smaller, each and this is easier to DIY into waterproof.
You also demonstrate that water comes into the innards easily via the propshaft and its tube. Yes, grease will reject it for some time, but at the scale of an Atlantic crossing? Human size adapted boats will have have a bilge pump that sends such water overboard. (6) I would consider an aquarium pump and check valve to send water out of the motor bay. Note that sea water is very salty and also very easily conduces electricity. With demineralised water in your espresso machine's steam boiler (never do that) its level meter might not detect water and fill infinitely. Check out recipes for sea water aquariums and you'll be astonished to see how many kg of salt you need for a 1,000 (1m^3 - not huge at, say, 0.6m deep * 0.9m high * 2m wide, filled to 0.83m high) tank. The trick in aquarium pumps is that they connect the drive shaft of the motor with the pump via magnets and there is a waterwall between pump and motor that way.
As it stands, the overall design is really elegant, but I would nevertheless (7) add a rudder. At the scale of the boat that may be sensitive (not sturdy) but it gives a backup plan for cases when one of the differential drives fail.
And, as the video demonstrates the inclination of the screws on the motor/propshaft coupling to get loose, I would (8) consider to fix those with Loctite when the boat is sent to the other side, or even glue parts together permanently. Parts of the Atlantic have weeds floating around (cf. Sargasso) and then there's human pollution - a sailor's nightmare when your propeller gets arrested by the "green" stuff or the pollution. Maybe in the software you can have a prop-cleaning routine that does quick bouts of alternating forward/backward spurt to try to loosen stuff from the props when the suspicion of that arises. (Some of today's digital cameras have actuators in their sensors to try to shake off dust. Does that work? Generally not, but often enough to want to have it.)
I *seriously* appreciate you taking the time to write this comment! With this current iteration being just proof that I can actually build something that floats and has working propulsion, I’ve skipped over a LOT.
I’m already looking to change the hull design to be self-righting, likely with a mono hull and center of mass below the waterline. For redundancy I’m thinking of combining the differential thrust with two rudders, and investigating how it handles navigation with eg one motor and one rudder available. Probably also investigate moving to using thruster style props, and completely avoid the prop shafts having to penetrate the hull.
You’ve raised some fantastic points, and quite a few that I hadn’t even thought about! I really like the idea of a prop cleaning routine, the compartmentalising of motors and electronics, plus handling ocean currents with “smarter” waypoints.
Definitely a long road ahead of development from this first prototype that’s for certain
@@akamatchstic - you're a great sport - excellent response to me back again. Note that in general the ocean currents are predictable (sailors use/have current maps), albeit there may be a tidal aspect to it, especially in coastal regions. That can seriously take a small vessel off course or prevent it from going where it wants/needs to be. That kind of relative consistency does not apply to what happens on the surface, driven by wind. If a sailor optimises their route, sets their waypoints and timing, then they take these things into account. This is only one tiny reason why an ocean going captain's education is an applied BSc (at least in my country).
With an electric vessel (EV) you're not sensitive to large regions of mid (E/W) and mid (N/S) Atlantic that can have zero window for long times each year, so you don't have to take that into account. But squalls can wreak havoc to the surface and if/then median waves can easily run up to 3m high. That should not be a problem as long as the waves stay sinusoidal in shape, but with high winds their tops fold over and can flip, or throw back, your EV.
So, sailors also make detours along current flows that are in their direction and they get to their destination faster than the shortest line would have given.
Thrusters are a nice idea that tries to solve an efficiency problem with propellers, but they add some resistance and drag. The "new" loop like propellers mimic the thruster tube without the tube and in the way a wing of an aeroplane causes lift, their shapes cause forward lift on top of the forward pushing helical motion through the water. These will become popular as fan blades too as they make much less noise (noise is lost energy and as these shapes loose less they make less noise).
The question what type of propulsion is best, is a hard one, though.
One challenge is to cut out transmission losses. If you can enclose the motors in a waterproof hydrodynamically shaped egg with the prop sticking out directly from the motor shaft, then that may be the most efficient. In your current model you have transmission losses because of the angle between the motor and propshaft.
Such a thing could be like an outboard motor where the motor is underwater. This is an optimisation problem again because in order to reduce transmission losses, you introduce an object with resistance. If, as a wild example, that motor were integrated into the ballast torpedo under the boat's keel then you have synergy.
In itself a thruster does not necessarily solve the leakage problem. Naval tech has dealt with that leakage since more than a century and propshafts may be lubricated in part by sea water.
Anyhow, pleased to meet you. Have fun.
It's 1m x 1m and quite deep. I have built-in enough volume and wide surface areas to be un-flippable.
When is the Atlantic Crossing? I have a almost built boat but need to install solar, pihawl. I built my catamaran out of wood and fiberglass.
No date set, I’ve got a lot more work ahead of me:
- hull redesign
- learning how to fibreglass properly
- electronics design with redundancy
- cooling system for the batteries and ESCs
- integration with iridium satellite network to send back gps updates
- solar panel research
- learn how to build my own battery packs
- investigate propellor design for efficiency (eg toroidal)
- investigate moving to thrusters attached to the hull and combining them with a rudder
kinda offtopic but i abolutely adore your hair
does he give you emma watson vibes or is it just me
Really cool project my man! Is the catamaran boat shape the final one for the Microtransat, or will you change the design for that one?
I’m thinking I’m gonna change it, mainly so I can achieve self-righting with the center of weight below the waterline. Probably go with a mono hull for it 😅
@@akamatchstic That sound like a good idea for the ocean waves. Looking forward to the update video, good luck!
I just need to install the electionics, I was going to use Pi, 4lte, I could put LoraWan and/iridium as a back up for small telemetry and course correction here and there?
Definitely would recommend looking at iridium, seems to be a solid way of getting data (even tiny amounts) to and from the craft!
Was worried if you had retired! I've been tinkering for months with a Jetson Nano since seeing one of your vids
How have you been getting on with it? Found that’s it quite a capable little system this end!
@@akamatchstic Never used Linux before, but think I'm over that hump. Needing to switch power to GPIO pins bc I'm switching FPC/HDMI/FPC for the cameras, though I probably should have just stuck with the ribbons.
bro please enable OpenPlotter with OpenCPN, olso i can't stream the internal raspberry py 4 camera into the mission planner camera divice . if you got a little bit of time, pls make a tutorial
make it so if a wave hits it, it righens itself
Try toroidal propellers for the efficiency
I’ll have to give them a try! Need to catch up on rctestflight’s video on the topic
Next video?
At some point it’ll be coming 😅 I’ve started work on some hull mounted thrusters to avoid water ingress in the prop shafts. I’m also in the process of moving house; right now my workspace is all packed up into boxes.
Dip your silicone tool in dishwashing liquid and it will work fine .. works with fingers too
I was trying straight water with no luck, will try this next - thanks!
@@akamatchstic might want to add some water if liquid is very concentrated...
A method to overcome water entering the hull could be to use a 'Z' drive which is common on tug boats "en.wikipedia.org/wiki/Z-drive"
Ghizzz crist! just get a cheap Ender V2 or larger chea FDM printers and avoid to much all that glue thing! the cost you've used in all that glue epoxy and all are just the same as a cheap big volume 3D printers anyways!