PDA

View Full Version : [Contest Entry] Project Big Bird



bradpowers
01-13-2008, 02:30 PM
Hi everyone!
Here's my submission for this contest, a (kind of) autonomous hovercraft. This project was originally for the final project of a mechatronics class, which I did with a group of four other students. We basically designed this to be a platform for further research into holonomic vehichles, artificial intelligence, and developing an algorithm for autonomous navigation. Here is a video of the hovercraft roaming around on our school soccer field:

http://www.youtube.com/watch?v=VuJnW5AvJik
This video is of the hovercraft under computer control via an Xbee Pro wireless chip, and a PIC microcontroller. We used the PIC to control one Novak brushless lift fan and two E-Flight Brushless motors with variable pitch propellors. The pitch is controlled by two servos which are also connected to the PIC. Also, sensor data is relayed from the PIC to the Xbee, which then is transmitted to the computer, which allows us to implement a SLAM algorithm, and control the hovercraft simultaneously.

We have a project website which I encourage you to visit, it has source code (at least the stuff that works well currently), hardware diagrams, design process info, and final specifications.
The website is: www.bradleyp.com/BigBird

Also, I have posted an Instructable on how to make the hovercraft, which is at
http://www.instructables.com/id/Autonomous-wirelessly-controlled-hovercraft/

Pictures of the craft are located on the project website, as well as in my gallery.

If you have any questions, feel free to post here or PM me.

Thank you!

Alex
01-15-2008, 02:28 PM
How easy is this to control? Does BigBird have any ability to carry objects? How much weight?


Also, sensor data is relayed from the PIC to the Xbee

What sort of sensors are you using?

Last question (for now, hehe), are you using a computer to control this, or what?


Neat project Brad (and cool project name too:))! I have your project submitted in our contest.

DresnerRobotics
01-15-2008, 02:41 PM
Nice team project! Mechatronics is the field of engineering I'll be working towards for my undergraduate.

Alex- I read through his project page when this was first posted, I believe they're using an ultrasonic sensor. And from the look of the gui, they appear to have some sort of remote control for use as well.

Matt
01-15-2008, 03:12 PM
This is so cool :) I used to dream of building one of these as a kid, only big enough to ride on. I think there was an ad in the back of boys life magazine or somewhere that was selling directions on how to make one out of a vacuum cleaner. I never got around to it, but that's probably a good thing since I doubt it worked, LOL.

I still think it would be cool to make though :) Maybe I can settle for following your project for now and putting a hamster on it.

bradpowers
01-15-2008, 03:54 PM
Alex - Controlling it takes a little getting used to, it's not exactly intuitive to drive a craft that deals with extremely low friction. After about five minutes of screwing around with it though, it's really easy to work. We have a "pocket" located at the front of the craft that is capable of holding objects, and we also have an acrylic plate that we can mount modules on to: for example, we mounted a video camera on there once, And we're working on other modules. I would estimate that the craft can carry five pounds above it's own weight (which is pretty impressive, as it weighs less than 5 pounds!).

We have code written for a number of sensors, but we didn't have the money to buy anything more than an ultrasonic sensor. I've written up a pretty extensive set of code for the Arduino, so that I could incorporate a GPS module, a 5 DOF IMU, and a few ultrasonic and/or IR sensors. At this point, it's just about being able to pay for those sensors (That's always the case, isn't it?).

As far as control goes, there are a few options. I can use a Futaba RC controller, with a switch to go between RC control, PC control, or a preprogrammed control. I am working on another program where I can set up waypoints and have the craft navigate to that waypoint.

Thank you for entering this project into the contest, and thank you for the complement. If you like the project name, check out our website, it's Sesame Street themed!



Tyberius - You'll love Mechatronics, it's a very exciting field, especially given the recent push toward making mechatronics easier to do for everyone (ie. the Arduino and such). I keep going back and forth between designing my own Mechatronics major and going with a robotics major.



Matt - I remember those ads in Boy's Life, I used to freak out every time I saw that. I've thought about making a rideable hovercraft, I just have to find the time. I would definitely give it a shot, making a hovercraft isn't so hard, and then you get to make it do stuff, which is the fun part. Ummm, hamster?????

Thanks everyone for the kind words, I'm glad you enjoy the project.

Dave
01-15-2008, 05:19 PM
Cool! I read the instructable for this when it was making the rounds on the tech blogs a few weeks ago. Making that autonomous will be a real challenge, from a control perspective. I have a feeling battling inertia is going to be a computational nightmare, and I'm looking forward to seeing how you handle that.

bradpowers
01-15-2008, 05:32 PM
Glad to hear that you saw my Instructable, that was pretty cool when it went around on Engadget and Makezine and Digg.

I'm worried about inertia too, I'm thinking that moving the thrust systems forward (I'm working on making two custom variable pitch, variable direction ducted fans, and I think if I move them closer to the center of gravity, I'll be able to avoid some of the weird torques and such. This whole project is kind of a computational nightmare, but I think it's doable.

Matt
01-15-2008, 07:57 PM
>>>I would estimate that the craft can carry five pounds above it's own weight (which is pretty impressive, as it weighs less than 5 pounds!).

I will second that. 5 pounds IS a big number. I was expecting you to say 1 pound or so.


>>>Matt - I remember those ads in Boy's Life, I used to freak out every time I saw that.

You do? That's hilarious! I have a vague recollection of a tiny ad with a picture of a seat on a triangular frame with a black disc at each point. It filled my young head with fantasies of "flying" around the neighborhood with it. I'd love to see the ad again, just for nostalgia.

>>>Ummm, hamster?????

well... I was running with the under 5 pounds thing :D

kdwyer
01-15-2008, 08:59 PM
Have you considered using some sort of larger rudder to help dampen/straighten movement?
And have you tried this (trusted this!) over water?

bradpowers
01-15-2008, 09:25 PM
Matt - Yeah, we were really impressed with the amount of lift that we were able to get. It's actually fairly simple though, our first prototype had about a third of the surface area on the bottom of the chassis, and could pick up about two pounds. It's reasonable to assume that increasing the surface area of the craft (to a point) should increase the amount you can lift. Also, the entire thing is made out of extremely lightweight materials.

kdwyer - Two things about the rudder. Despite how it looks in the video (ie. completely uncontrollable) that is definitely not the case, it is very stable, I was just having a little fun out on the soccer field! Also, we did some research that suggests that rudders are horribly inefficient on hovercraft, for example, to make the thrust be in a direction five degrees of the heading of the craft, you have to turn the rudder to 45 degrees. Because of this, we decided to not use a rudder.

As for the water thing, well, the pond that we would use to test seaworthiness was frozen over, and the pool staff at the college next door to us wasn't so happy about a hovercraft in their pool. I honestly believe that it would be safe to run on water, as the chassis is made of foam that floats, so if we found a waterproof enclosure for the electronics (the lipo batteries!) we would probably be fine. We shouldn't have any problem getting to hump speed (the speed at which the craft really gets moving over water) with our propulsion fans, they were only running at 10%

Tymtravler
01-16-2008, 08:38 AM
>>>You do? That's hilarious! I have a vague recollection of a tiny ad with a picture of a seat on a triangular frame with a black disc at each point. It filled my young head with fantasies of "flying" around the neighborhood with it. I'd love to see the ad again, just for nostalgia.

:D


Hey Matt I can take you back in time, check this out.

http://www.hovercraft-aircar.com/

Tymtravler

Alex
01-16-2008, 09:10 AM
I just checked out your site. The about us page with comparison's to Sesame Street characters is a riot!

Seriously though, the mechanical design (http://www.bradleyp.com/BigBird/Mechanical_Design.html) page was very interesting. I figured that you just started with the square "traditional" design, but I see I was far mistaken. I was pretty impressed that take 2 worked at all! Did it hover steadily, or did it sort of just bounce around in the blue foam disk? It was such a simple design.

Great work on the Software (http://www.bradleyp.com/BigBird/Software.html) page as well! I love the yellow Big Bird GUI.

I'm really out of my realm in the next few sentences, so please bear with me a bit;) Do PIC's have the ability to communicate with serial devices? The reason I ask is because I was a little curious if there was any specific reason(s) (besides the obvious of it being much easier) why you chose to use the analog output with the EZ-1 instead of the digital serial output. I'm not claiming that I know a ton about this topic, but I just read #9 on MaxSonars FAQ (http://www.maxbotix.com/MaxSonar-EZ1__FAQ.html) and it sounds like using the analog output has more potential of introducing errors?

Alex
01-16-2008, 09:11 AM
Ha! That is priceless TymTravler! That sounded familiar when they were talking about it, but I couldn't picture it in my head till now, thanks:)

Matt
01-16-2008, 11:13 AM
That's awesome, what a great find. I also never knew that it was such a well known ad. There was a link right to an actual website for the "Air-Car"

http://www.hovercraft-aircar.com/default.htm

It's like the exact web equivalent of the print ad. All promises and no proof. If it works then where's the freaking youtube video of it?

Anyhow, I wasn't trying to hijack the thread. Back to our regular programming.

bradpowers
01-16-2008, 11:34 AM
Alex - Yeah, the mechanical design was pretty... interesting. We started out wanting to be able to control pitch and roll on the craft, and have the craft be able to recover from a cushion failure, so we went with the circle design, and decided we would have five of those little hovering disks all at the same time. There were a few problems with this. The biggest issue is that we didn't really know the right things about hovercraft at the time. For example, we got 10 free computer fans (big beefy 120mm computer fans) that had extremely high flow rates, but we had major issues with air coming back up through the fan, and basically nothing happened, until I turned up the voltage to the fans to about 20 V (waaaaaaaaaaay out of spec!) and fried them. Also, the thing just looked too good (*cough* heavy *cough*), made out of aluminum and all. Lastly, when you make a skirt for a circular craft, it doesn't work to just cut out a rectangle of fabric and try and get it to play nice. The skirt gets all wrinkly, and leaks out of the bottom of the craft, causing it to just sit there.

With design number 1 out of the way, we thought about what was wrong with the craft. We couldn't really pinpoint what exactly was wrong with the craft, and we were about a week behind where we wanted to be (pretty sad when you have a month to work with!) so we (the MechE's) split up (the ECE's were busy trying to not fry PICs). We had two basic ideas to work with, the blue foam square, and the more conventional chassis. Given our preliminary results with the blue foam (yes they float, but they can't lift a paperclip, much less a battery, or circuits...), I decided to go against the grain and make a more conventional chassis (I'll dig up the video and post it soon) out of hot glue and corrugated plastic and tape, and tape, and a little bit of sketchy thrown in. This used the brushless lift fan that we used in the final version, and worked! I called over my teammates, and we put together a more conventional skirt, and the thing just fired right up and floated. At this point, we had wireless control of the world's sketchiest hovercraft, which was what we called our "minimum deliverable" basically, if we got to this point, we considered ourselves victorious. From there, we figured out that since we have this great source of underskirt pressure (the significantly overpowered lift fan) we could simply increase the size of the craft and get a linear return in lift capacity. That's what you can see now.

As for the ultrasonic sensor, yes PICs can communicate over serial, but I think that we had used up our serial communications working with the XBee, but I'm not sure, so I emailed someone who will know and I'll get back to you.

bradpowers
01-16-2008, 11:35 AM
Tymtravler - Nice

Azinman
03-26-2009, 05:28 PM
Hey When i was in 7th grade i remember a kid who took a leaf blower and stuck the nozzel into a hole in a 4'x4' piece of 1/4" scrapwood. He the took a tarp and strategically punched holes in th bottom for the air to flow through, thus making a working hovercraft! We all gotta take turns riding it it was so cool :D