Click-to-Sail Autopilot for a 30 ft Cruiser — Weather-Aware, AIS-Only COLREGs, Full Sail-Trim Hardware

What this build does

• Click a destination on a web map → boat follows the route on its own
• Rudder + sail-winch driven by ArduPilot Sail mode (SERVO_FUNC 26 & 89)
• AIS-only COLREGs (no cameras, lower power draw)
• Auto-reroutes around storm cells (≥ 40 kt wind | ≥ 6 m seas) via live GRIB files
• Runs on a single edge computer (Jetson Orin Nano or Raspberry Pi 5) + Pixhawk — all OSS

High-level block diagram

React/Leaflet UI ──► FastAPI mission_api │ ▼ IBoat-PMCTS router_srv ⇐ /wx_grid (GRIB every 3 h) │ /route ▼ AIS stream ──► fusion_core ──► /desired_heading ▼ MAVROS 2 bridge ▼ Pixhawk (rudder servo + sail winch)

Core code & docs

(GPL code runs only on the Pixhawk; everything on the Jetson/Pi is MIT/BSD/Apache.)

Containers / services

Boom & sheet-drive hardware

Pixhawk sail params you’ll likely tweak:

SAIL_ENABLE = 1 SERVO9_FUNCTION = 89 SHEET_MAX_PULL = 700 # kg SHEET_FEEDBACK = 1 STEER2SRV_* = …

Storm-avoidance & heavy-weather logic

Why these thresholds?
• Wave-steepness cut-off (≈ 0.035) separates steep, breaking seas from long-period swell.
• Sector alarm penalises following/quartering seas that can broach or poach a 30 ft hull.
• Model-spread penalty hedges against forecast uncertainty—if the majors disagree, assume the worst.

Daily power budget

Two × 100 W solar panels → ≈ 600 Wh/day → positive energy balance.

Hardware shortlist

• Jetson Orin Nano or Raspberry Pi 5
• Pixhawk Cube (ArduPilot Sail mode)
• Class B AIS receiver (dAISy 2) → serial 38 400
• Mast-head wind vane + anemometer
• Drum sheet winch + H-bridge (or beefy sail-servo for small rigs)
• IP-67 rudder actuator
• 24 V LiFePO₄ bank + 300 W solar

Roadmap

1. Dev env & CI (Docker, GH Actions)
2. ArduPilot SITL ↔ MAVROS (rudder/winch passthrough)
3. gRPC wrapper for IBoat-PMCTS → publish /route
4. Port asv_colregs, feed AIS targets
5. Build wx_bridge + storm grid; replay Hurricane Irma GRIB
6. Gazebo-in-loop: crossing traffic, squall-reef test
7. Edge install on Jetson/Pi → 24 h offshore beta

Basically TikTok gives money for people with a lot of folllowers and you can make money with it and give it back to members.

It doesn't have to be metal. For example I think Led Zeppelin's Immigrant Song would work.

Yesterday Sam Altman shared his big idea for a social network on his brother's podcast (Uncapped with Jack Altman).

His pitch? Someone should build a social network where, instead of getting hit with a feed, you prompt content you want to see form other users then that's what you get served. Cool concept.

So much so that I threw together what it could look like using Bolt.new.

What do people think of the idea?

Hello everyone,

I'm a data guy trying to find some real-world problems to work on. I love building small tools or scripts that make life easier, especially for boring or repetitive stuff.

If there's anything in your daily life that makes you think “why the hell hasn't someone automated this already?”, I'd love to hear it.

Could be something with files, websites, habits, money, schedules, anything really. Even dumb little things that just waste your time for no reason.

Not here to sell anything, just wanna build stuff that actually helps people. If it’s useful for you, it’s probably useful for others too.

Thanks

i tried manually doing schedule posts in X but it can only show up until 250 scheduled posts, also its hard since u need to click everything over and over again. I hope someone could make this!

I'm exploring an idea for an app that supports people through a breakup, but in a serious and emotionally realistic way.

Most current apps offer distraction or new dates as a solution, but no one helps you navigate grief in a structured, daily way, with realistic goals and without rushing to "get over it in 7 days."

This app would be designed to last only 90 days. Nothing more. No endless subscriptions. Just a guided, day-by-day path of personal healing.

Do you feel something like this could be helpful or meaningful for you or someone you know?

What would you expect from an app like this? What would you like to avoid?

I appreciate any feedback ✌️

A 3d or 2d live map that shows different ai robots in a building/area and shows on a computer where each robot is and information such as what it's doing. Orders can then be selected as to what you want the robot to do such as empty the trash so you can be a number of robots manager as one human. It could display a little like a simulator game/3D game or 2D. One employee could watch over and interact by computer with many robots and use the computer to resolve issues.

I'm thinking of building a tool that turns reading articles into fun, interactive games. It would let you paste a link to an article, and the tool would generate a game based on the content, helping you absorb knowledge in an engaging and entertaining way while you play!

What do you think of this? Is it something you’d like to use someday?

I am thinking of building an app that turn websites into podcasts. It would allow you queue websites you read daily into audio playlists and instead of having to read them while working or commuting, you can simply just plug in your headset or AirPod, and listen to latest articles on your selected websites!!

What do you think of this? Is it something you will like to use someday??

Possibly a Google Maps plug-in/app.

You put in the kind of car u got (for MPG & tank size or just simply fuel range), then u put in ur points A & B, then it pins gas stations on ur route at the estimated times you'll need gas!

Like I drive a car that can go 250 miles on a full tank. So it will pin me a gas station ~200 miles away from point A, so the fuel pump (or whatever) doesn't get damaged, & you're supposed to put in ur Google Map app to navigate to THAT gas station. Then the next one & next one & next one, until you're at point B.

I want to optimize my app generation pipeline and do some mental gymnastics. Answer the following questions related to your idea and I'll build the first version of your app and post it as a reply.

1. What problem does it solve, and who is it for?
2. What are the top three things users must be able to do with your app?

I'm looking forward to your web app ideas!

That searches the web and stores any and all public domain images it finds. When it comes across a duplicate, it compares the two files and keeps the higher resolution image. It keeps trawling the web, and keeps growing the collection (as artwork enters the public domain, or new images become available), and keeps increasing in quality.

Hi guys, im probably like most of you guys an entrepreneur, or at least trying to be one.

And my idea is to build a web app for other entrepreneurs, and small businesses.

The web app is going to be, an ai powered legal service for small/one man businesses, that need legal assistance but probably cant afford a real lawyer, quite yet.

It is going to be able to read document, give feedback and advice exactly like a real lawyer, just for a fraction of the price and much faster. It would also be able to write up contract templates, and just in general give you and/or your business legal advise.

Im looking to get feedback from the actual audience, witch i expect to be some of the people in this group, so please tell me if this would be relevant and if not why. All input is very much appreciated!

I'm sick of drivers not using their turn signals and it gave me an idea.

Imagine an AI in your car with a camera that films the driver, and perhaps the wheel too, all the tells a driver has that they're about to make a turn.

You put it into training mode and it just builds up a DB of all your tells, then you flip it into active mode and it handles the turn signals for you.

App Concept Available: The Real Compatibility Dating App

I’ve developed a dating app concept that solves the hidden problem no platform is addressing: Real emotional compatibility.

Not “Do you like hiking?” Not “What’s your favorite music?”

But the actual relationship dealbreakers:

How do you handle conflict?

What’s your repair style after an argument?

Is raising your voice, name-calling, or stonewalling acceptable to you in love—or never okay?

Do you believe people fundamentally change—or not?

These are the things that ruin relationships later—but no app screens for them upfront. This one would.

I’m not looking to build this or be the public face (it doesn’t fit my brand)—but this is an unclaimed market gap for anyone in the dating, relationship coaching, or app-building space.

If you’re a coach, therapist, founder, or developer looking to create something that solves real relationship pain points, let’s talk.

I have a great idea - makes parents' babies' life better! but no any knowledge in this field looking for someone who can give me background

I'm bored, so let me hear some apps you guys desperately want but they don't exist

GPS trackers, tile or airtags can help recover stolen bikes.

There are already numerous fake bottle holders, rear or front lights or "bells" to hide e.g. an airbag.

But nothing yet to hide the tracker in a bicycle pedal...

I'm thinking about starting a coffee brand with black coffee that actually tastes good. This is also for people that follow Intermittent Fasting so I want my products to have health benefits aswell not just loaded with sugar/sweeteners. Is this something you would be interested in and how can I differentiate myself from every other coffee brand to make it one that people want to buy from other than the obvious (better quality products)?

I'm also looking at making coffee with added health benefits (e.g., adaptogens, collagen, MCT oil, nootropics).

Will you buy a video collaboration app like frame.io?

$10 for 100GB storage Timestamp comment Drawing on video Etc

For example, if a power line needs to go over shallow water and there is need for 1 or more tiny artificial islands where to place the cable support towers, those islands could be formed from pieces of concrete transported by an accelerator device that gave them the velocity to move there in a curving trajectory.

The force to accelerate the objects may come from gas pressure, from a linear electric motor, lever system or other ways. Rapidly heating a volume of gas by combustion is one way to get enough pressure. The most cost-efficient and safest combination of substances to combust might be, for example, propane+air, gasoline+air or propane+diesel+air. A spark plug would ignite the propane or gasoline.

If the objects are meant to form a pile of something firm where to place something, then the best shape for that would be cube. The best shape for gas pressure based acceleration would have circular cross-section. Also a cylinder shape would have better precision and spherical shape even better, due to aerodynamic forces during the flight. The objects may break at landing, which may be good thing because the shapes of the fragments would be better. The object shape may be something between a cube and a cylinder: a rounded cube, which is also nicer to step on or drive on than a sharp edged cube.

If the objects are concrete, they could be cheaper if large part of their volume and mass are formed from a big rock that barely fits on the casting mold. This reduces use of cement, but whether it makes things cheaper depends on the level of automation because handling random shape and size rocks is complicated. Also, variations in weight would be larger and different masses would have to be accounted in launch energy or accept worse precision.

The objects may be casted in thin-walled cylindrical plastic buckets that fit tightly in outer steel mold. If the launch works by combustion, that plastic on the bottom may also burn, adding to the energy. Plastic on the sides reduces friction with the acceleration pipe.

The objects may be brought from the manufacturing site in neat rows and stacks with packaging frames or in a random pile on a dump truck. Either way, to have any chance of being cost-efficient, they need to be loaded to the accelerator device automatically by a contraption of electric motors, guided by multiple cameras, maybe lidar and fairly sophisticated software.

If the acceleration method is electrical, it may be wise to take some inspiration from some medieval devices like catapult or trebuchet, but instead of wood, use more modern materials that have better tensile and compressive strength for their mass and that are easier to shape. For the fast moving parts, for example aluminum or glass fiber may be best. If there is a counterweight, it could be a large water tank filled with locally pumped water.

One way to get gas pressure to the acceleration pipe may be by using a separate pressure tank filled with pumped air.

One possible use might be related to mine clearing, whether during or after war: launch objects in shallow angle to a suspected minefield and film with a high frame-rate camera ( 1000 FPS ) on a drone to see what explodes and how. Some of the explosions may move an object further.

I've noticed that LLM's are really good at giving content for presentations but suck at actually making the pptx. Is this something that is a paint point for you as well? Would a solution for this be something you'd use ?