Society of Robots - Robot Forum

General Misc => Misc => Topic started by: PuddleJumper on October 11, 2007, 11:41:25 AM

Title: Design Phase: Aquatic Project with Multiple Parameters - Looking for Advice
Post by: PuddleJumper on October 11, 2007, 11:41:25 AM
Ahoy!

I have recently finished meeting with a group of friends and we believe we have finished what we think we want conceptually.  If I seem a bit vague in some areas it may or not be intentional.

The concept is this: Design a Lead Bot and 5 Drone Bots; each bot should fully waterproof and able to perform a set of given actions detailed below.   


LEAD BOT:
#1 - Remain Stationary except by RC command
#2 - Send 3 Commands to Drone BOTS
        A. Spread out from LEAD BOT
        B. Begin Programmed Actions
        C. Converge on LEAD BOT
#3 - Send Beacon Signal to DRONE BOTS to allow them to stay within a designated range.
#4 - Send commands to DRONES based on RC inputs from USER

DRONE BOT:
#1 - Spread away from LEAD BOT
#2 - Converge on LEAD BOT
#3 - Move in a random predefined motions (For example Move N would be a predefined motion, and wiggle could be another)
#4 - Accept Signal from LEAD BOT for RANGE and COMMANDS
#5 - Have an RC override

OTHER USEFUL PARAMETERS:

* DRONE could sense when it was trapped by or collided with a potentially dangers item (not just running into a chute of grass
* DRONE will need a range of approximately 25 yards from LEAD, but up to 50 would be ideal
* DRONE should also keep a safety range of 1 yard from other drones.
* LEAD and DRONE must be 100% water proof, and should be fairly resistant to minor bumps and bruises.
* DRONE and LEAD chassis must be relatively small - 8" x 4" x 3" or less would be ideal
* DRONE and LEAD should use metal props for movement
* DRONE and LEAD should move at speeds of 1-3 feet per second.  Variable speed would be nice.
* Chassis should be weighted enough to allow for some stability in rougher water (read lake on windy day, not ocean)

Please feel free to ask more questions if you are curious about function and form.  As of right now, I am a bit Leary of giving out the use (nothing illegal, or unethical).

I am looking for advice across the board - from materials to sensors to motors.  Any past experiences? Right now, we are leaning towards a plastic chassis with a lead keel for stability, low frequency RC transmitters, a mercury switch for one of the predefined motions (tipping into a dive position), rechargeable Nihm batteries (need suggestions), and two small RC boat DC motors for the propollesion. 

I am new to robots, have only built one and read two books.  I can use any help you are willing to give.
Thanks any Advance,

PJ
Title: Re: Design Phase: Aquatic Project with Multiple Parameters - Looking for Advice
Post by: amakusa on October 11, 2007, 01:26:59 PM
definitely read the tutorial on waterproofing servo motors, bluetooth has a range of 25 to 35 ft I think wireless has a range of 33-100 yards, not sure what the under water range would be though. Variable speed for your motor can be achieved by a microcontroller, you could use IR if they are close together, although visibility in the water will effect its range. The rest of the spreading, converging and RC override can be programed. Your chose of material to use depends if you are going to use a swim bladder or not, if so you may want to use a material that matches the density of the water, and above all else talk to JonHyland I believe he is a member of this forum he has built a AUV already and would be a wealth of info to you
Title: Re: Design Phase: Aquatic Project with Multiple Parameters - Looking for Advice
Post by: PuddleJumper on October 11, 2007, 01:45:03 PM

I just finished watching the waterproofing video segment - I must say I am uber impressed by the wealth of knowledge, not only in that post, but in the over 100 post I have read this morning (my brain hurts).  I will check out the HUV website as well, and hopefully he'll catch this post as well.


You post brought to light some things that I failed to mention in my original post.

#1.  The BOT will have two parts: 1 being the on the surface "shell'.  In the shell will be several things.  The shell will contain a couple actuators to perform various activities (not yet decided upon)

#2.  The BOT could infact house the "brain" and "eyes"  and "mouth" in it's shell. 

#3.  The BOT's second part will be it's "drive train" (for lack of a better term).  Underwater should be the propeller.

#4.  A key feature of this BOT (the DRONE's anyway) will be to keep the "Drive Train and Chassis" and a horizontal level to the water, and have the shell lift to a 90 degree angle from that.  (Picture a boat pointing it's nose to the bottom of a pool).  I realized that this is gonna take some experimenting with weight and center of gravity, etc.

Well, thanks for taking a look... I am back to research mode. 

Title: Re: Design Phase: Aquatic Project with Multiple Parameters - Looking for Advice
Post by: amakusa on October 11, 2007, 02:10:30 PM
so like two boat put together? one on top of the other? 90 degree- I'm picturing a sail? drive train? you should keep it simple with just a motor, maybe you should try experiments with an r/c boat?
Title: Re: Design Phase: Aquatic Project with Multiple Parameters - Looking for Advice
Post by: PuddleJumper on October 11, 2007, 02:45:35 PM
Two pictures... see if this helps.   ???

Title: Re: Design Phase: Aquatic Project with Multiple Parameters - Looking for Advice
Post by: amakusa on October 11, 2007, 02:59:52 PM
OK next question, why this configuration? does it serve a practical purpose? and with it off set like that you will need the weight of a motor to balance or at least some kind of weight to balance it out, also if they are just going to float PVC is the best way to go
Title: Re: Design Phase: Aquatic Project with Multiple Parameters - Looking for Advice
Post by: PuddleJumper on October 11, 2007, 03:41:29 PM
Great Questions!

#1.   The off set config. is "I think" necessary for one of the primary functions of the DRONE.

Lets say I was using a laser gun to hit a target on the shell.  When that target is hit with the laser, I want the Shell part to tip up and dive into the water, while the counterweight to this would be the motor weight, and possibly a metal keel.  I just starting drawing it out, and I am thinking of using the battery (if it is non-bouyant as the front weight).  Basically some sort of mechanically device will have to release and draw back the shell to it's original position.

This function is a "BIG BONUS TICKET" - if I had to scrap it, I could; but would be disappointed.

#2.  Yeah, should just float.  I was looking at PVC, but then read the tutorial on plastic HDPE, and figured I would check out that as an alternative.


Just starting doing some thinking on the RC aspect of this.    Lets say I have 6 BOTS on the water (1 LEAD, and 5 DRONES).  And DRONE#4 starts going somewhere that while within the PARAMETERS, it's not somewhere I want it.  I need to be able to move DRONE#4, but not move all the drones or the LEAD DRONE.  Is this gonna be doable or am I getting into a massive problem area?  Further more... Let's say I have 18 BOTS on the WATER (3 LEADS, and 15 DRONES) Now I sense I am getting into a large problem.

My tentative hope is that it would be possible to 1 RC control talk to 3 LEADS and have them command their DRONES?!

In a nut shell... In the 18 BOT scenario.  I would have LEAD-A, LEAD-B, LEAD-C and DRONES(A1,A2...B3,B4...C5)... I would want to flip a switch on the RC remote to A, B, or C - to command the given LEAD, then maybe another switch 1-5 to tell it which DRONE(1-5) to command.

Looking for the simplest (and cheapest) way to accomplish this...  Hope ADMIN doesn't boot me! I know this should probably be in the other folder, but the DESIGN PHASE seemed to fit best here.

THANKS,

PJ
Title: Re: Design Phase: Aquatic Project with Multiple Parameters - Looking for Advice
Post by: amakusa on October 11, 2007, 05:58:31 PM
flipping up could be accomplished with a spring latch design with a winch to bring it back down, a laptop programmed correctly would allow you to chose the boat you want to control, HDPE is a good alt.
Title: Re: Design Phase: Aquatic Project with Multiple Parameters - Looking for Advice
Post by: Admin on October 11, 2007, 06:06:13 PM
The biggest issues you will have that I want to point out are:
waterproofing - dont underestimate it - a tiny leak will fry everything!
wireless communication - this is crazy hard underwater
sensors - its really hard to see underwater . . .

for communication, consider acoustic modems . . .

or another idea, have each of your drones operate on a timer, and when the time is up have them all surface to communicate, then submerge again with new mission plans.
Title: Re: Design Phase: Aquatic Project with Multiple Parameters - Looking for Advice
Post by: amakusa on October 11, 2007, 06:09:19 PM
The biggest issues you will have that I want to point out are:
waterproofing - don't underestimate it - a tiny leak will fry everything!
wireless communication - this is crazy hard underwater
sensors - its really hard to see underwater . . .

for communication, consider acoustic modems . . .

or another idea, have each of your drones operate on a timer, and when the time is up have them all surface to communicate, then submerge again with new mission plans.

acoustic modems why didn't I think of that DOH!! actually admin I think there boats not submarines at least thats the impression I get
Title: Re: Design Phase: Aquatic Project with Multiple Parameters - Looking for Advice
Post by: PuddleJumper on October 11, 2007, 07:01:53 PM

Quick Update:

Looking over the 'specs'  - I believe that all sensors (input and output) can be located above water. Don't want to try and transmit underwater.  I think somewhere on this site I read the KISS theory.

As for as controlling all the BOTS out on the water - I would like to apply KISS theory there as well... I was reading on another thread that I could "program" (not sure that is the right term) the OUTPUT on an IR beacon to send out different signals.  How diverse are these signals?

For example could I have the LEAD BOT send out a IR = 310? 310 might tell BOT#3 to STOP, 410 tell BOT#3 stop...  (Given microcontrollers my guess is the number would actually have to be under 256 - is that correct?) Bas

So I could set the R/C remote to have a switch sent A-E (for the 1 to 6 LEAD DUCKS), and 1-6 (for 1-6 DRONES).  I set my Switches to A1 and hit the STOP BUTTON.

This would send LEAD BOT A a signal to send out the IR signal 110 (from example above) and BOT#1 would stop. 

Am I just completely off here?  I guess one of my main problems is not understanding what language R/C Remotes transmit in.  And what language the receivers transfer to the micro processor.  I read the Tutorial on R/C remotes, but it didn't go that in depth.  Anyone have a link to that?

NEXT STEP:  Determining layout of IR sensors to make sure LEAD bot can send info out and DRONES can get it in.

FOLLOW QUESTION: Anyone have any arguments for BLUETOOTH in this type of application? If so, any links to info?


THANKS,

PJ
Title: Re: Design Phase: Aquatic Project with Multiple Parameters - Looking for Advice
Post by: amakusa on October 11, 2007, 07:52:44 PM
Q: What is the range of Bluetooth transmitter/receivers?
A: Bluetooth is designed for very low power use, and the transmission range will only be 10m, about 30ft. High-powered Bluetooth devices will enable ranges up to 100m (300ft). Considering the design philosophy behind Bluetooth, even the 10m range is adequate for the purposes Bluetooth is intended for. Later versions of the Bluetooth spec may allow longer ranges.

http://www.mobileinfo.com/Bluetooth/FAQ.htm#g1 (http://www.mobileinfo.com/Bluetooth/FAQ.htm#g1)

tracking system for leads and boats so they know where each other are
Title: Re: Design Phase: Aquatic Project with Multiple Parameters - Looking for Advice
Post by: PuddleJumper on October 11, 2007, 09:03:26 PM

First I must say that I am very impressed with the wealth of knowledge on this site.  I pretty much used 5 hours of work, and now the last few hours at home reading everything I could on this site.  Kudos!  I only hope I will be able to provide some help to others as you all have helped me.

After researching R/C control, I am leaning towards using it across the board.   I think it possible, but is gonna be program intensive (which I don't mind - I happen to have a C programmer in my back pocket)  Let me put this out there the best I can.

I know that old garage doors openers used the 8 pin DIP switch (I hope that is the right term) to set which door opener it talk to.  So my idea is this:

#1  Place a number of transmitters in the LEAD BOT each with it's own DIP switch.  Each DRONE will have a receiver with a DIP switch set to match 1 of the DIP's in the LEAD. 

#2.  Purchase a higher-end R/C controller that can handle sending the following commands to each LEAD: 1) ACTIVATE ALL DRONES, 2) SUMMON ALL DRONES, 3) DRONE-X DUAL MOTOR (Basically using tank like steering to move - forward, back, left, right, spin)

* Where DRONE-X is any one of the variable drones individually.

This solves several problems, but I am having trouble figuring out if I can program the CONTROLLER to talk to multiple LEAD BOTS.  Can the same type of DIP switch type isolation be used? Am I gonna end up building my own controller?

NOW FOR NEXT PROBLEM: THE TRACKING SYSTEM

Keeping my DRONES within a given range of the LEAD.  I am trying to get around IR technology given the outdoor, poor weather, sunny weather, bumpy water, low emitting width, poor receiving, environment.  And also because placing the IR emitters/receivers on the shell might alter the shape of the SHELL in a none acceptable manner.   Is there any type of beacon / homing device that will could be used here?  Keep in might there might be 3 LEAD BOTS on the water at the same time so need to make sure each DRONE knows who it's LEAD BOT is and vice-versa.

THANKS,

PJ
Title: Re: Design Phase: Aquatic Project with Multiple Parameters - Looking for Advice
Post by: PuddleJumper on October 11, 2007, 09:20:05 PM
Just found this link:  Not much of the "HOW" but I have a feeling this is the type of technology I am gonna have to put into the DRONES and BOTS.  Might have to buy one and reverse engineer it.  It looks like it uses Bluetooth (or 2.45 Ghz ISM) and some kind of FM frequency.  Anyone know anything about it?

http://www.loc8tor.com/introducing-loc8tor.asp

Not to mention this is a pretty cool item if you are absent minded like me.

Cheers,

PJ
Title: Re: Design Phase: Aquatic Project with Multiple Parameters - Looking for Advice
Post by: Admin on October 12, 2007, 04:50:39 AM
Quote
Anyone know anything about it?

http://www.loc8tor.com/introducing-loc8tor.asp
Someone has asked a similar question already but I couldnt find the post. The problem with this is that it doesnt give you direction to the target, only distance away.

As for transmitting to different drones . . . instead, just transmit on the same frequency to all the drones at the same time, but have the signal be something like a list, where each drone will only 'listen' to the part assigned to it.
Title: Re: Design Phase: Aquatic Project with Multiple Parameters - Looking for Advice
Post by: PuddleJumper on October 12, 2007, 07:37:45 AM
Thanks! I wasn't sure if you could send out a message from the LEAD say 128, and just tell all the drones but one not to listen to 128.  That makes sense.

That locater will tell direction (or so it says) as the handheld device claims to have an LED display that will tell you left, right, forward, backward, or even lower or higher in elevation as well as distance.  I assume if the handheld (or locater) can display this information then the data must be gathered somewhere.

At it again for another day,

PJ
Title: Re: Design Phase: Aquatic Project with Multiple Parameters - Looking for Advice
Post by: JonHylands on October 12, 2007, 07:38:39 AM
I would strongly advise you to look into building a network of devices using xbee wireless. You basically broadcast messages to all devices on the network, and the packets have addresses to determine which device the message is for.

Each device on the network has a unique ID, and each device ignores any messages not meant for it. You typically want to use a master-slave system with an open network like this, so there are not any traffic collisions. You can also have a special ID that represents a true broadcast, which all devices will process (but not respond to).

The trick with master slave is you have one device on the network that is the master, and the rest are slaves. Only the master is allowed to initiate communications, and the slaves it communicates with must respond only when asked.

- Jon
Title: Re: Design Phase: Aquatic Project with Multiple Parameters - Looking for Advice
Post by: PuddleJumper on October 12, 2007, 10:48:14 AM
Interesting!

I have been reading up on the XBee tech.  I has good ranger (300') on the smallest unit XBee? ZigBee OEM RF Module.

If you have used this tech before I have a couple questions:

#1. If this was an off the shelf type product (say you buy a group of 6, 1 master 5 slaves) how could it be programmed so that you could be sure you never had two of the same masters if you bought to groups?

#2.  If I were to buy a single slave - how could it tell it who it's master was, and tell the master who to command?

#3.  Would the CONTROLLER work on the same type of tech?  Like a small PDA type handheld?

OTHER THOUGHTS:

I have yet to mention costs as an issue.  And truly it isn't too much of a concern, but I would like to create a group of 6 for less than $3,000 for the first run.  Hopefully if I moved into producing more groups I would eventually like to get the cost down to around $240-480 for a set of 6 - This shouldn't include the CONTROLLER which I imagine is gonna cost a bit more.

Am I out of my mind?  I see all the little parts running around 8-$20 bucks each.  I know the shell is around $10.  I fear that one part could blow me out of the water if I don't choose the right technology.

Just looking for a bit of input so I can start to budget out my BOTS.

Thanks again for all the advice,

PJ




Title: Re: Design Phase: Aquatic Project with Multiple Parameters - Looking for Advice
Post by: Admin on October 12, 2007, 11:29:14 AM
For Zigbee, have you checked the manual? (I havnt read it, but Im sure its in there)

As for costs, I recommend making up an excel sheet, entering each part and detailed cost to calculate the total. Dont forget to include taxes/shipping and mistake corrections. I do this for all my projects so I can budget it appropriately.

And time is money too!
Title: Re: Design Phase: Aquatic Project with Multiple Parameters - Looking for Advice
Post by: JonHylands on October 12, 2007, 11:34:29 AM
The difference between master and slave isn't a hardware thing, its a software slave. The master code is programmed onto one robot (or maybe your handheld remote control), and all the other ones are slaves.

How you go about actually doing this depends entirely on what kind of computer you put into each device.

- Jon
Title: Re: Design Phase: Aquatic Project with Multiple Parameters - Looking for Advice
Post by: PuddleJumper on October 14, 2007, 11:22:16 AM
Ok, so I love the wireless idea.  That's now a must.

I am thinking a servo for a "turret-like" item on the SHELL.  Is it true these can stop at a given point?, If so I am gonna use it for two small flaps on the SHELL, and possibly for the TIP-UP operating.

So that would be 4 Servos, and 2 DC Motors (for propolsion and steering).

If we include the 6 devices, the XBEE (Looking at the XBEE STANDARD) for Remote Control, and then some type of tracking (I am still looking into Loc8tor); What am I looking at as for as Controllers? I have read the tutorial and am still a bit hazy on how to figure out requirements.

Thanks,

PJ