Small Boat Autopilot With Obstacle Override

You should start by researching aircraft autopilot systems. That would help your understanding of the problem space.

Aircraft use TCAS (Traffic Collision Avoidance System) for collision avoidance. That system would not be practical for boats (unless a significant number of autopilot boats exist) as it requires every craft to have a transponder broadcasting position, altitude, speed, and trajectory. However, some lessons might be learned from terrain following radar or ground proximity warning systems.

first i would like to thank you for your feedback

the main problem for the aircraft model is that all thesis and books i have read assume that i have a huge background of autopilot modeling which is not true

for the obstacle avoidance system, the main goal is that in many times the control wireless signal is lost, so the boat is lost

i wanted to make a system that derives the boat back to me in such case

in its way back it might hit a ship or a floating barrel and cause it to go down or even the other ships propellers may pull the boat and destroy it

the boat is less than 4 meters long only!!!!

I'll be with you next time. GA

Hello hanyarnaoot,

i think acoustic sensor will be good

Why do you think this? In a marine environment, especially on the small, robotic scale you seem to be talking about, I think there would be an extreme amount of interference. What is the actual size of the boat, and what types of bodies of water will it be operated in?

Mike

first i would like to thank you for your comment

the boat is to be used to deliver small spare parts and such stuff between ship it is less than 4 meter (12 feet's) long

the boat may hit another ship or a floating empty barrel or a big piece of wood and so on, which may damage the boat

if at least i managed to determine that there is a ship at a certain direction (by the propeller noise) i will avoid it but how(marine radar cost is high and in the same time the boat is too small and short for one)

if there is any thing in your mind that efficient and low cost please tell me

thank you very much

There are marine radar units designed for smaller boats. Although the cost may seem to be high right now but in the long run you may save a lot more by using it.

Noise, not only from the ships props but also the water itself is going to be a huge problem to over come, if acoustics are your choice of detection.

Also, you will need to detect the roll of the boat, as this will inter fear with the auto pilot. Roll will cause the auto pilot to over correct.

Ultrasonic range finders would only be good for very close distance detection. ( I tried this for several years )

The problem I always had, was when the boat would roll it would pickup the water as an object. I tried to mount the sensor within a gimbal it helped but it was not good enough.

I ultimately wound up spending the money on a radar unit (Ratheon)

Good Luck !

Dave

With regard to OPs #2).

Why didn't you buffer and average the ultrasound/acoustic data ... and reject any data with less than a 50% hit. This should take care of the waves, if you use a long enough sample time. If the vessel is not a fast mover, it should be appropriate to use a sample size of 3-5 seconds. You could always tweak the hit rate and sample length in real world, if not in real time. IE ... when a 'hit' is detected, lower the sample time and make the hit detection more sensitive .

With that said, an acoustic sensor should be all you need ... I can't see a 4 meter craft used as a cargo vessel used in more than half a meter seas ... otherwise you'd be looking at a sub ... until it hit bottom ... :o0

In fact, a semi-submersible with a tall tower might be a good idea ... depending on the regs in the area of use. (That) should minimize the roll and yaw.

Unfortunately, most maritime regs that I know, prevent unmanned over-the-horizon vessels.

It would be less expensive and time consuming to put rubber bumpers around the boat (or foam ones.) and also to have a reference point guidance system. By your statement, that the boat is for delivery, that suggests the distance is not too great. I would recommend using buoys with radio beacons, and using a standard signal-strength scan and pid loop to steer by. If the distance is only a few hundred yards (like a river), steer by rope and eyelets.

but i do commend your ambition. I think you will have to be committed to learning, and have the time to play with it for .. (years)

chris

If I am reading you right, you are shipping parts from ship to ship. Are these ships at sea, and what type of distances are there between these ships?

I am wondering based upon your response I hope you provide if there is a much simpler solution.

By the way, has anyone thought of the military implications of such a craft?

Sounds like a torpedo. I wonder what the cargo is. Seems like a fine way to go blowing up oil rigs.

Yes x 4 - hence 'no comment'

& GA

Let us not declare war until we are at war. That is the tool of the fear mongers. If we stopped all discussions of technology with military applications, CR4 would be vacant.

I agree it is appropriate to have some sense of the person we share with, and would like to see more of the OP responding... but otherwise, I think we should play it straight.

Terrorism will never be stopped by advanced nations withholding technology. It will be stopped by advanced nations stopping the dissemination of war materiel to warlords, and it will be stopped by advanced nations giving a shit about the poor underpriveleged people of the third world countries, and making strides to help feed, shelter, clothe, and water them, etc etc etc..

/rant off.

Chris

Caution is the better part of valor

a new member with no real body of work to refer to, asking a question that has some potential for mischief.

can make one reluctant to fire up the "A" game

original posters have to meet certain minimum requirements to get meaningful responses. there have been several new members recently that seemed to have questionable motivations for joining our little goat rodeo.

it's up to our new friend to provide more context for the solution he seeks.

agreed..

_{^{(It is your fault if I'm cranky today anyway...)}}

_^chris

thank you for your understanding

please read my reply

thanks again

I can't find any other reply from you?

see number 4

thx

sorry i thought i have posted it

please read my last reply that is posted on reply to #10

i am a newbie to this site

so do not expect much from me

Fear is your friend. I have suspicions about the OPs motives. Why boat so small? Why concern with passive sound aquizition of "target"?

CR4 has published interesting cheap tech, even into 100 dollar RC video drones.

Not much of a profile, from OP. I've not changed mine in awhile, but if you look you will find I want to see people from outerspace on the 3D Space TV, and prevent Apocalyptic Riot through airport islands of commerce.

Not sailed much but even in the intercoastals of FLL a 14 foot long motorboat is hard to handle. 26 foot sailboat is biggest I've steered into port. Out it was 6 to 12 foot swells. Below deck I had to stop drinking after two beers and get some air.

Regardless of all wonders realized and available to all, terrorism will continue because it is part of human nature for some humans to hate themselves and others for nothing but their humanity.

Some people simply will never be happy unless they are unhappy, and spreading it around makes them happy to be unhappy.

first if all i am so sad for what you have said

my intention are far away from what you have said

and i want to tell you something

an enemy of the US may read your comment and decide to make an attack thanks to you (remember that the idea of what happened in 11/9 was taken from a movie)

also those people use a different, autopilot than what we are discussing they are the suicidal boomers ( USS COOL was destroyed by one)

second

using it to destroy an oil rig is none sense because small boat like this will not go that far to be in a range of a one

third

it is not a torpedo such small boats will not go faster than 10 knots, otherwise it will cape size

nearly all military ships in the world exceed 30 knots easily (how come it will be on third of its speed)

I must clear things much more for every one

I have recently got my master degree in automatic control

but i was very theoretical and i did not feel that i have learned a lot

for those who are specialized in such stuff it was about the relation between system state controllability-observability and system state sensitivity to change in system parameter (quiet boring is not it)

my professor told me that there is a PHD scholarship but there is a large number of candidate for it

so i wanted to come up with something real big and astonishing

in the same time applicable and will be a very good additive to my CV to work as a professor after my PhD

the idea came to me because a close friend had a boat renting facility for diving and such stuff

and in many case the rented boat my suffer from a mechanical problem an need a new spare part like spark plugs,fuel hoses.....even sometimes fuel when the renter does not estimate the distance enough, food, water ,wine, extra diving gear

so he had to keep one boat from 4 not rented fro such cases

so a huge boat is used to deliver a fuel hose of 1 feet long

so i talked to myself and said a small boat that is to carry a cargo of not more than 100 lib. will be fine

but my friend is too fat

and really the boat may not withstand him

so it came up in my mind why do not i make such a boat to go to the designated area where my friend boat is located to deliver the needed spare parts,wine,water,fuel......

and my friend liked it very much and told me he will fund the project within the range of one of his boat price (he will not pay to the boat operator)

(by the way it could not be used in smuggling because it will not withstand open sea state)

and people Will be amazed because it is a huge project

and most important of all i will learn much about applicable automatic control

and get a free scholarship , become a professor

in brief life will be so good

has any one any objection on that

thanks to any one who tried to help me

Welcome to CR4

home of the cranky old guy

Your project makes much more sense with the additional information.

Tell us more about the parameters of the project

budget

range

power source

....

You could consider investing in this and tie it into your gps,

http://www.systron.com/products/motionpak%C2%AE

As for a collision sensor a seperate ultrasonic system would seem to be one way to go.

Here is another link which could be of intrest.

http://www.ocean-server.com/compass.html?gclid=CIGvr4SRk6QCFQ25bwodUwsuHg

Not to expensive.

"Here is another link which could be of intrest." (interest)

"Not to expensive." (too)

Well caught , such observation. here is a link for you.

http://www.worldwidewords.org/qa/qa-nit1.htm

But equally; I never seem to find them when I don't read and think about a post.

Agreed: under many (I think "most") conditions, vision will be far better than sonic/ultrasonic sensing. It will have less noise - but still plenty - and automatically increases the importance of nearby objects over far-away ones due to apparent size in the view. If you can combine vision with direction input, you don't need the actual range: you need to find targets, and see whether their direction relative to yours changes. The old saying is, "Constant bearing means collision." So if the direction is changing, ignore that target. If not, change your own direction or speed to make it happen from your own end.

I also agree. I use vision systems to detect failed part in many forms. They work well once set up. Excellent suggestion!

Rumba Boat. Taps something, remembers it, goes around it.

Do you mean Roomba? If not, what DO you mean?

If you do mean Roomba, how do the situations correspond: the vacuum cleaner works in a relatively unchanging environment and repeats its path, or tries, but the boat is in a constantly-changing situation and may never be called upon to repeat a path. One has three degrees of freedom, and the other has six - which is sort of why this thread got started, if I understand correctly.

Use your eyes.

Don't sleep on such a boat, unless it is docked or on the land. If you want to eliminate collisions, look out and steer it by hand. Once I thought I was on a rescue mission in the North Sea to rescue a Jet Ski with 2 persons on board.

Why?

Because they left 3 hours earlier, with only 90 minutes fuel.

So I took a small rubber dinghy with a 5 Hp outboard and went into the ocean.

After 10 minutes, I spotted them, moored against a boat, full with boobs and booties. To show me what an idiot I was, they came to splash around my gummi boat, but instead of avoiding me, the pilot maneuvered the Jet Ski straight onto my dinghy.

Probably with a speed of 40 Mph plus. No damage to me and/or the dinghy

A flying passenger that took a launch of 50 meter through the air and made a very rough landing on the water. The pilot only had to get 5 stitches, nicely divided under the lower lip.

This proves the idea of a Canadian Fellow poster.

The first rule on the ocean or on the waters that you share is to not be a danger for the other users. You need to know the laws of the water.

Your first issue: Correcting displacements automatically with such a small boat: forget it unless you are very fast and can afford lots of wasted fuel. Also the coastguard may consider you (a)drunk.

You need to anticipate the incoming waves on beforehand, or you might find yourself upside down from time to time. So be PRO Active, not Re active.

To not sound to negative today:

you could work with some infrared cameras, or put a night vision scope (waterproof with wipers) on your nose. They work pretty well with fog too.

Or install some tweeters on the boat and make them auto balancing (horizontally) and, send some high audio frequencies away from you in different directions and listen to the reflecting sound.

In fog, use a bell or fog horn- others are supposed to do that also.

Then in the dark you should use navigation lights and also watch for them on other vessels. They give you the size and direction to anticipate on too.

If there is a lot of dangerous wreck stuff, make a deviator or deflector in the water on your boat (with shock ab- or adsorber) You can push them or detect them or even catch and collect them.

A standard GPS helm should work on calm seas, even on a floating motorized bath, bed and beyond. Listen to Tina Turner songs at home.

thank you Garthhthe

the boat itself which is most expensive component is old fiber glass boat with an electric motor that was used by children to play with in nearby water

range is less than 2 miles (including the working instrument power usage, go back journey and 20% of battery power as a reserve)

the main problem is that the power to weight ratio (for batteries) is low

and by mounting additional batteries the cargo decreases dramatically

does any one know a better electrical power source than ordinary batteries

i was thinking of fuel cells

but i do not know about the cost, space needed, power to weight ratio

on the other hand changing the electric motor with a fuel powered motor will cost a fortune (motor price, hull modification, redistribute weights to maintain boat stability -that why i do not wish to put a sensor that requires a tall mast like radars-) and also the the vibration may be another problem

budget is about 40 grant (i wish to make it as small as possible)

by the way no one asked me about the way with which the boat will work

how would i get it to work

i thought of a radio on each of my friend ship to tell him what they need and their current location, so my friend will put what they want on my boat and enter the location and required way points to the boat computer and the boat will go on its way

but there is a problem

if the boat is heading for a limited maneuverability ship the people onboard that ship must have some sort of short range remote control to guide it to the ship because

1-even if a GPS on L2 frequency or a differential GPS are used there still a position error for the main boat or the small boat

2-as far as i know all available autopilots keeps you in a circle (usually 1 to 3 times the boat length) not in a point, no one managed to do better than that

3- the main ship itself may change its position a little bit for any reason

Remote control making is not a big deal

i will just connect the receiver unit on the small boat to a transistor or a relay to pass a signal to the moving motor

by the way the most suitable stuff for are the satellite moving motors

and if any one wants i could post how to connect them to rudder and speed regulator arm

and how to get a feed back of their position to a computer and how to initiate motion through computer joystick and a simple circuit mounted to computer parallel port