Saturday, January 31, 2015

Basic Hardware Plan

The basic hardware plan. The solar panel connects to the charge controller, and that to the battery. The 12V USB converter makes 5 volt for the phone, relay controller and maybe a bluetooth speaker. The relay controller controls electromagnet, horn, lights, and motors. The power for the things the relay controller controls comes from the 12 volt battery and those wires are not shown to simplify the drawing. The phone talks by bluetooth to the relay controller and speaker.

The electromagnet(s) will be used to disconnect from a towed boat and maybe to deploy a water brake. However, the current plan is for the water brake to just drag a little bit and deploy automatically if the boat starts to go backward.

The relay controller really uses up 2 USB slots.

Note that the phone also has its own battery and could keep transmitting GPS coordinates through the night (Internet or SMS) even if the boat were resting on the water brake and the main battery was dead.

We also plan to have the phone controlling a GoPro camera, which is not shown in this drawing. For a long time we will probably just put the GoPro on the boat or a towed boat and have it shoot video without any control by the phone. The long term plan is to use some camera, either on the phone or GoPro via wifi, to look at lights on other boat for automatic docking while at sea.

Thursday, January 29, 2015

Comparison to Other Fun Solar Robot Boats

There are a few other student/hobby projects to make ocean going solar robot boats documented on the web.

Scout

The Scout Transatlantic is a 12 foot long solar robot boat. It used a trolling motor and a rudder. It made it over 1000 miles into the Atlantic.

Scout uses two Arduino Mega microcontrollers; one to handle navigation, and one to handle sensors and communication. Some electronics for Scout below:

FishPi

The FishPi is also a solar robot boat. FishPi uses a Raspberry Pi. Also using a trolling motor and rudder. It seems a lot of electronics work was done to get GPS, compass, controllers all working. Not sure it was ever tested in the ocean.

Piribot

The Piribot uses Arduino controller and Android Phone. It is 1.12 meters long and has a big keel. Their goal is to cross the Mediterranean north of Alexandria. This is more than 1000 km but not as rough as the Atlantic. Were estimating June 2015.

Thoughts on how we differ from previous solar robot boats

In our design we are using an Android phone with compass, gyro, GPS, communications, and camera, and not developing any new electronics. Since the only software we have to write is for the phone, our software development environment seems much better than what these projects had, so it will be easier for us to get the software working.

We are working at a smaller scale, like 1/10th the weight of scout.   This means we can make a model out of styrofoam and cover it with styrospray. This is far easier than what they had to do to build Scout.

For the weight of our boat we have a very stable design. Waves will not spin it left and right and we think it will take waves larger than usually seen in the Caribbean to tip it over. The models we have so far are probably not ready for the Atlantic. If we build a bit bigger model and put a battery inside each float we could make a self righting model. Then I think we could cope with the Atlantic in the right time of year and right path.

We are using 2 motors and steering by turning motors on and off, with no rudder.

Since we are using little motors we can easily put on 4 motors. This would let us keep working even if one failed.

We will probably have some sort of water brake (tested version 0.1 already). This will help the boat stay in place if we have to rescue it or stay in place overnight after the battery runs out.

The other boats went for satellite link. For some time we will stay near islands in the Caribbean where roaming by our phone is probably good enough. May eventually get a sat link.

If you spend years making one boat and it is lost at sea, that probably kills the project. Our hardware will be so easy to build that as long as we have money it won't be too painful to build another version if we lose one. Our main investment will be in the software which will not be lost if a boat is lost. :-)

Our work is leading toward a human scale solar boat of similar design (quadmaran). So software development and experience with this type of design is working toward a future project.

Also, we are want to be able to have boats dock together while in the ocean. So one boat could rescue another.   Also working toward human scale boats that could join together in the ocean.

Other Related Stuff

The wave powered wave glider has some solar for electronics but not propulsion.

Underwater gliders that use change in boyancy to glid forward while going up or down. These are underwater and not solar powered, but interesting.

There are a number of sail powered robot boats that use solar for electronics. For example, ASV Roboat. There is even a World Robotic Sailing Championship. It seems a students at the University of British Columbia have won this competition several times and are now planning to send a robot sailboat across the Atlantic.

There is a Solar Challenge - World cup for solar powered boats but these are with human, not robot, drivers.

Some Robot Boat Parts


Click to enlarge and see parts better

The solar panel and battery were each about 2 lbs and online said 2.5 lbs. So we are under our weight budget so far!

Online the 12 volt battery said 10-14 Ah Pb Eq so I was thinking we were near 150 watt-hours. However, up close it also says 51 watt-hours. So I now think it is like 1/3rd the capacity I was expecting. Could not do much motoring during the night with this. The water brake might be needed for overnight. Night trips are some time off yet, so I won't worry too much about that for now.

The lights are really bright. The horn is really loud. The motor has very good torque.

Still don't have the phone, relays, or propellers. Will probably get these next week.

Also not sure yet on how the motor, stuffing box, and propellers are all going to work mechanically.

Meet the island boys / Water brake version 0.1

Meet the Cate crew
Arranging us from shortest to tallest, our initials spell out our last name
Corey
Amoni
Teryn
Ethan

Our Dad and the boat and water brake

The boat and water brake in the Ocean

Getting into deeper water so the water brake does not get scratched on the reef

It filled with air and stayed above water

But with a little persuasion the bag stayed underwater

We shortened the water brake so it was on average a little more forward and it worked a little better

Corey (the kid in the blue hat) on a glass bottom boat

Conclusions:

Flapping plastic has lots of air drag. Ideally water brake is all under water.
Water brake has to be far forward to keep boat pointed into the wind
Don't want a bag that can fill with air or water
Wood floats and so is not ideal for something that is supposed to go down into the water.
Does not need such a large water brake

Wednesday, January 28, 2015

Hardware Design Choices

Friends on facebook have asked a couple questions about why I chose the hardware I did so I thought I would repeat /and expand on the answers here.

Android Phone vs Arduino or Raspberry Pi

The kids and I had already written some apps using App Inventor and were very impressed with how easy it makes software development. In the "wifi connect" mode as you change things on your screen your phone is running the new version. If you want to really install it you just click "build" and have your phone scan the 2D barcode on your screen and it will install. It is a graphical environment that is great for my 11 and 13 year old boys to learn programming in.

Arduino and Raspberry Pi are better for doing custom electronics hardware but not nearly as good on the software side. In the end the things I want to do will take a lot of software but we don't need custom electronics. The phone can control a bunch of relays and has all the sensors and communications we need (GPS, Gyro, Compass, Acceleration, Internet, SMS, wifi, bluetooth). The phone having a display makes debugging the software much easier. The phone also has its own battery, so even if all else is down it may still be able to send messages. You can get a phone with a case for $85 and you don't have to put it together. The memory, CPU, and communications bandwidth of an $85 phone is far greater than what was used to send men to the moon. So we should be ok with the phone.

Brushed vs Brushless Motors

I did a calculation with a prop with a 1.4 inch pitch and decided that 3000 RPM is what I wanted. With the solar at 12 volts and the battery at 12 volts, the motor had to be 12 volts. For the 50 watt solar to be able to charge up the battery during the day while running the motors and for the battery to just about make it through the night the motors could only be around the 7 watts of this motor. I only want to get things from Amazon so it is easy for someone else to copy us. With these 4 constraints, this motor was the only one we could find, so I did not have to choose between brushed and brushelss. :-)

Variable Speed Motor Controller vs Just Relay

With a single part with 8 relays we can control both motors, some lights, and electromagnets. We can do a lot of stuff with this. By turning relays off and on we can still turn the boat and reduce power usage at night. It would be nice to have a variable speed motor controller, and to be able to do reverse, but I don't think we need it. Also, I could not find a bluetooth motor controller on Amazon and did not want it bad enough to order it from someplace else.

There is some chance that we will later decide that automatic docking without reverse is just not practical; however, I think we can get a sea anchor/brake to work and with one boat using a brake the other will be able to dock with forward thrust only. In the Caribbean we almost always have the trade winds that are the scale equivalent of hurricane winds for these small boats. So I think we can count on the wind to push back if they just stop pushing forward.

Weight Limit

We tested the model with sandwich bags full of sand and think that we want try for a weight limit of 6 to 8 lbs. The solar is 2.5 lbs. The battery will be either 1.2 or 2.5 lbs. The two motors together are about 1 lb. So we might make it. If we run over we could add flotation to an existing model or build another with more capacity.

Cost

I would like to have the possibility of the boat going on its own for days. But realistically we may lose a boat or two before this is all done, so keeping costs down is important. It would be fun if other people built boats that were software compatible and this seems more likely to happen if costs are kept low. The first set of parts was bought with Christmas money from grandparents.

My Goals

I want my boys to experience the thrill of doing some cool engineering, learn about programming, test out this design, and develop software that could be used on large solar boats of similar design later.

Relay documentation

Our relay still is not here but we found the documentation for the relay online.

Codes to send Relay controller

Left motor on=¨e¨, off=¨o¨
Right moter on=¨f¨, off=¨p¨
Light 1 on=¨g¨, off=¨q¨
Light 2 on=¨h¨, off=¨r¨
Magnet on=¨i¨, off=¨s¨

Higher level movement codes

Forward=¨ef¨ - both motors on
Right=¨ep¨ - left motor on right motor off
Left=¨fo¨ - right motor on left motor off
Stop=¨op¨ - both motors off

Tuesday, January 27, 2015

Status Plans Ideas Goals Fun - Project Index

Already finished

App using GPS
App storing to web and reading stuff from web on another phone
Compass app
Bluetooth between phones
Phone talking to bluetooth relay to turn stuff on/off.
Tested water brake
Figure out if relays or electromagnets mess up compass
Put electronics on boat and waterproof
Figure out motor/driveshaft/stuffing box/propeller details
Use phone as bluetooth remote to test hardware
Log GPS,bat level, status, to website so easy to see how/where boat is
Navigation control - Adjust heading to keep track on course for waypoints.
Point to a gps location and flash lights with minimal forward motion
Send Morse code in flashing lights
Keep track of maximum tilt to see how close to tipping over.
Got phone charging off main battery via USB charger

Future

Tow second boat and release it using electromagnet
Use how fast it can turn boat with one motor on to estimate battery level (may get bluetooth volt meter if not happy with this method but motor speed is proportional to voltage so it should work)
Go to a nearby island and back (1/4 mile).
Go to one location then take a picture aimed at a certain location or direction (ocean shots of houses for real estate).
Put 4 motors/props on a boat for more thrust and fault tolerance
Reducing power usage at night if needed.
Downwind costing mode. Just use minimal power to keep pointed in the right direction.
Get pictures from phone working reliably (works many times now but can crash)
While on sea brake record/report wind direction
Mode for motoring toward the wind within limits. If it shuts down it will then probably slowly drift on sea anchor right back towards you with lots of time for rescue.
Phone to make gopro take pictures at top of each wave.
Algorithm to explore the edge of phone service around island
Use microphone on phone to figure rpm on motors.
Use microphone and spinning thing to measure wind speed
Dock two boats with person using phone as bluetooth remote control. One boat with sea brake on and other docks by slowly controlling how wind pushes it back into that one.
Dock boats using software control with colored lights at different angles to make it easy to get right path like airplane landing lights.
Have phone send texts to several phones when in trouble
Drive in big circles in sandy hill bay for more than 24 hours.
Go to France and back, St Martin is about 7 miles :-) and test roaming data
Go along near reef with GoPro under water making movie
Blow horn if indication a bird landed (funny acceleration, tipping, slow boat)
Get robot boat to catch fish (motors/props by front floats and fishing line by back).
Have one robot boat able to rescue another robot boat with broken motor(s)
Convoy of robot boats that can change order if front one breaks
Trip around Anguilla
Longer and longer trips - multiple days
Obstacle avoidance, mostly floating seaweed patches
Cross the Caribbean and back
Satellite data link
Cross the Atlantic

This list will be updated with new ideas and plans. The top things in the "future" section are probably being worked on or will be soon. This page will link to future posts from items here. This post will be the index for the project.

Monday, January 26, 2015

Robot Boat Project Parts

We've ordered the parts below to convert the "Boat" above into a autonomous robot boat. The basic idea is the android phone can talk to the relay controller over blue-tooth, and the relay controller can turn off and on all kind of things. The phone that we've ordered has a compass, gyro, gps, camera and internet! We figure the computer in the phone is more power than the computers they used to go to the moon. We have started programming with app inventor.

300G.cm DC 12V 0.58A Brushed Electric Motor 3000RPM

Two of these.

Performix Clear Liquid Tape Spray

11LB DC12V Holding Electromagnet Lift Solenoid / Electric Lifting Magnet

8 Channel Smartphone Relay Bluetooth Remote Control Kit

Only need one relay controller. Thought we wanted 8 but that does not run on USB power and figured out that 4 will work so went with that.

4 Channel Smartphone Relay Bluetooth Remote Control Kit

Estone 1 Set 3 Blade Nylon Propeller P1.4 Dia 47mm 52mm 55mm 57mm For 3/16 Shaft RC Boat

Battery Tender BTL09A120C Lithium Iron Phosphate Battery

There are larger batteries but to save weight in the Quadmaran-5, and for while just operating in the daytime, we moved to the smallest one. The quadmaran-10 will use more and larger batteries.