PC based autonomous boat

Overview:
This project is in the very early stage of development. I've got a couple ideas that I want to explore and need a suitable platform to work with. The following are some of the components I hope to use to assemble this platform.

Objective:
The objective of this project is to design a very small boat ( roughly 1 meter LOA ) that will be capable of autonomously traveling a preset survey grid to produce a sonar bottom scan. For this first iteration of the project I hope to use all off the shelf components, the only customization will be some custom code I will need to write.

Components:
CPU
HP Mini 311 - The brains of the boat. Probably over kill for the task, but I happen to have one.

Operating System
Windows XP - Mostly because of the wide range of free tools and info available for it.

Control Software
RoboRealm - This very cool software is really a key element to the whole system. It supports a wide variety of hardware (servo controllers, GPS sensors, serial interface, etc.) and provides a simple interface for making them work together.

Motor / Servo Control
Parallax USB Servo Controller - This will be used to control the boat's propulsion motor(s) and rudder(s).

Sensors
GlobalSat BU-353 USB GPS - The GPS will be used primarily to provide position, course to steer and distance to waypoint. It will not be used for heading information as GPS heading is almost completely worthless at slow vessel speeds.

Silicon Labs USB tilt compensated compass - A very inexpensive, tilt compensated, USB interfaced compass that I just happen to have.

Route planning
Polar View - A free chart viewer, route planning application. A track can be marked on a chart and then saved as an ASCII CSV file, which will then be used by the autopilot software I hope to write.

Remote monitoring / Control
Wifi + VNC - My plan is to use a 1 watt router connected to my shore based computer, and then a 1 watt USB dongle network adapter on the afloat computer. The goal will be to be able to monitor the boat's position and sensor data as well as being able to update/change the autopilot's route. I have no idea what the potential range is for Wifi over water, but hopefully 1 watt will give me at least a couple hundred meters.

Boat
I'm still sorting this out. I'd prefer to use something off the shelf (ie. a large RC boat) but most don't have the weight carrying and stability capabilities I will need. The hull will likely be a catamaran or trimaran and propulsion will be by electric motor. The final system will need to be hand carried by one person, so the size will likely stay at roughly 1 meter LOA.

I've been working to do something similar the last few months. I have a fly-in fishing lodge in Canada and I'm working to map our lakes and create contour maps for our guests to use. I've been mapping the water for the last couple of years using our fishing boats and various Lowrance GPS chartplotters that record all of the required parameters. I've written some Matlab routines to more easily visualize, edit and plot the data. I've just purchased Dr. Depth, but haven't used it much. It looks like it will make some nice maps, but we'll see how well it will work with our 7,000 acres of water.

Since I don't have the time to map the lake with any spatial resolution, I've been working on an autopilot for one of our boats. I'm using a Make Controller to interface with the Lowrance via serial, a linear actuator to control the tiller on an 8 Hp motor, and a very old boat. Obviously, safety is the most important part of the project, so it'll have some routines running to check the position and kill the motor if needed.

As a side note, I often wonder why I don't see more projects using the make controller. Just having the ability to write in C makes it worthwhile for me. Its obviously a bit big and heavy for a small aircraft, but for other vehicles I think it could be useful.

You need to be a member of diydrones to add comments!

Join diydrones

Comments

Shad July 8, 2010 at 7:01am

@bigkahuna - I'm using one of these linear actuators. They're $140, but have the power and throw needed for my application. This model has a built in pot so you can sense the position. I have all of the hardware and can control everything from the Make Controller, but haven't had time to code the autopilot yet. I also have a rotary actuator (from a junk car window) that I may hook up to the throttle to better facilitate launch and retrieval. Of course, I'll be wiring into the kill switch as well. We're pretty busy, so I haven't had much time to work on it recently.

I agree that it will be easier to debug in the boat. The actuator isn't that fast that I couldn't take control in the event that I needed to. Not much can happen at low to moderate speeds anyway.
bigkahuna July 8, 2010 at 12:18am

@Shad - You've given me another idea. I have a 12 foot inflatable with outboard I could use as a test platform. I wouldn't use it un-manned, but could use the system to automatically drive a sonar scanning track. Would be great to find a heavy duty linear actuator with an RC control interface, what are you using?
Shad July 7, 2010 at 9:13pm

I talked to some Humminbird engineers a couple of years ago and they said that there wasn't a way to extract the lat/lon/depth from the sidescan data. Does anyone know if that's still the case? If you could extract data from the entire width of the sidescan, it would greatly reduce the amount of passes needed for a good map.
Bill Porter July 7, 2010 at 7:56pm

Lol, I am using the Humminbird side scan sonar 1197c in a project at work right now. We are not pulling raw data from them, just a video signal of what the bottom looks like. We removed the Sonar board from inside the head unit and threw out the rest.
Shad July 7, 2010 at 6:49pm

I've been working to do something similar the last few months. I have a fly-in fishing lodge in Canada and I'm working to map our lakes and create contour maps for our guests to use. I've been mapping the water for the last couple of years using our fishing boats and various Lowrance GPS chartplotters that record all of the required parameters. I've written some Matlab routines to more easily visualize, edit and plot the data. I've just purchased Dr. Depth, but haven't used it much. It looks like it will make some nice maps, but we'll see how well it will work with our 7,000 acres of water.

Since I don't have the time to map the lake with any spatial resolution, I've been working on an autopilot for one of our boats. I'm using a Make Controller to interface with the Lowrance via serial, a linear actuator to control the tiller on an 8 Hp motor, and a very old boat. Obviously, safety is the most important part of the project, so it'll have some routines running to check the position and kill the motor if needed.

As a side note, I often wonder why I don't see more projects using the make controller. Just having the ability to write in C makes it worthwhile for me. Its obviously a bit big and heavy for a small aircraft, but for other vehicles I think it could be useful.
bigkahuna July 7, 2010 at 3:52pm

@Harald - I'd agree with everything you've listed except #2. As I'm not a C++ programmer and I've used almost all these components before with RoboRealm, I'm certain this will be the easiest way to go... at least for this first generation test.

@Sgt Ric - This is just a test platform to experiment with. My last project was an underwater ROV that was rated to 250' depth, so I'm well aware of the affects of water on electronics. I may never put this first model in the water, but if I do it will go in a Pelican case or something equally water tight.

@Mogly - I've got the Parallax USB SSC and it's supported by RoboRealm so that's what I'll use. I also have a Pulolu Micro Serial Servo Controller which I might use later if I decide to go with RS232 instead of USB. I've used both in other projects and haven't had any issues with either one.
Moderator

Sgt Ric July 7, 2010 at 3:13pm

As the guys are pointing out...just like most R/C planes WILL crash sooner or later, ...your boat WILL take on water sooner or later.
Harald Molle July 7, 2010 at 2:32pm

My advice:
skip the PC and use an Ardupilot instead!
1. Much cheaper
2. Much easier for the integration of RC equipment and GPS
3. Less weight
4. An Ardupilot may survive beeing submerged, a PC definitely not :-)
5. The Arduino environment is very very easy to start up.
6. don´t know...
Mogly (Umesh Tharanath) July 7, 2010 at 1:26pm

I would Suggest to use SSC-32 Servo Controller instead of Parallax USB Servo Controller...
Xander July 7, 2010 at 1:07pm

While the PC route may seem cheaper right now (considering you already have many of the components), think about what will happen the first time that boat takes on water. Swapping out an arduino will cost a lot less than swapping out a mini pc. With UAVs its often best to keep as much of the expensive equipment on the ground as possible. Crashes suck, so I really make an effort to keep onboard components cheap. Same would probably apply to boats.

of 3

This reply was deleted.