Arizona State University Autonomous UAV Project

[erichq2]

Hey everyone,

My name is Eric and I'm the student team lead of a group of mostly aerospace engineering majors at Arizona State University. We're part of a bigger organization here that is basically giving us a small or decent budget (depending on how you want to look at it) to do something educational and STEM related so the group has decided that we want to build an autonomous UAV for the AUVSI UAS competition in June. None of us have experience with remote control planes but we have experience designing planes and laying fiberglass so we're hoping to be able to build our own from the ground up.

So the first step is determining what we need our plane to do. Well, we need autonomous waypoint navigation, failsafes and a safety link, a twenty minute flight time minimum, and of course an imaging system.

We will be using Paparazzi as our autopilot because it is low cost (well it's free, but the hardware) and fulfills our needs. Paparazzi is a good starting point for us because we can't afford to buy commercial autopilot software with our budget and I think programming our own functional autopilot software would be an entire project on its own. Maybe in the future we can look into this if we recruit a computer science major, but we're all aero majors and we'd rather see our end product as a flying plane rather than a couple thousand lines of code. We'll purchase the board soon and when we get it, we'll take one of our member's rc plane and throw it in there to see if we can get a plane to fly autonomously by December 1st.

The drawback to using Paparazzi is that it uses IR sensors to determine attitude which I don't think is a very accurate and good way to go, but we have been graciously donated an IMU by our faculty mentor so one of the major projects will be to figure out how to make the IMU work with paparazzi.

Our imaging system will most likely be a really crude system to begin with because we're starting from scratch. Most likely we'll have a video stream and a filter to spot discolorations and a digital camera that will be controlled manually so that we can take a few still images when an operator spots a target/discoloration. In the future I'd like to improve on this but for now we need a plane that can fly.

As far as the hardware goes, we're still deciding on that. This plane will most likely be in the 10-20 lb range, so at first we threw out electric motors but now we're reconsidering. I've been looking into this myself and it seems like electric motors are more trouble than they are worth, that and I'm not sure if the motors would create noise that would interfere with our electronics.

The airframe, like I said, we'd like to design and build our own airframe. We have experience with writing sizing codes for airplanes in our classes and it would be "fun" to be able to actually apply what we've been learning for the past three and four years. However, if we decide that it would be best to purchase a frame, then I would not object either.

Right now we're a bit in a rut because we can't do much without knowing how heavy our plane will be. We need to decide on gas vs electric for a motor, and which ever one we pick, we will need to decide on either how much battery or gas we need and in the electric case, what prop will be needed to draw enough current to fly for 20 minutes. We also do not know how to determine what servos we need. Bigger servo = more torque but how much torque do we need? The main thing we need to find out is an estimate of how heavy our plane will be sans motor and battery/gas.

Anyways, I made this post here because it seems like you guys all know a thing or two about remote controlled planes and UAVs, so comments, suggestions or information would be nice :).

[Gary Mortimer]

I might well have a look here if I were you http://diydrones.ning.com/

There are very many clever people here that will fill you in on all the other stuff.

A trip down to your local RC flying field might be a good place to start finding out about airframes and having a crack at flying your own.

Also find a good simulator to play with, there are a few free ones about.

Best of luck

G

PS I find it amazing that a group of aeronautic students dont have an RC nerd amongst them!!!

[AntonK]

I would love to see some proof that IR sensors are not "accurate" Maybe Im biased as Ive been a Paparazzi developer for 4 years, but I can tell you that the University of Arizona's MAV program along with ENAC University has flown circles around schools trying to use a autopilot with an IMU for years now. If you want help integrating an IMU with paparazzi let me know I have one in my helicopter.

AJ

[erichq2]

I would love to see some proof that IR sensors are not "accurate" Maybe Im biased as Ive been a Paparazzi developer for 4 years, but I can tell you that the University of Arizona's MAV program along with ENAC University has flown circles around schools trying to use a autopilot with an IMU for years now. If you want help integrating an IMU with paparazzi let me know I have one in my helicopter.

AJ

I was not aware that UA had a UAV program. Rivalry aside, it sounds interesting and I am always up for reading about what other schools have done especially if you have been working with paparazzi for four years. I'm not sure what IMU he is giving us but any help integrating an IMU with paparazzi would be much appreciated. My only "proof" is our faculty adviser who did his MS and PhD work in helicopter UAVs. He mentioned that he has worked with paparazzi before but disliked the IR sensors for various reasons.

I'm actually waiting for a friend to come pick me up to go to Tucson for the weekend right now so maybe we could talk about this in person if you'd like?

[danstah]

The real problem with IR sensor is they are not calibratable in a real precise way. They have to be tuned by using your eye mostly rather a measuring device. They are however plenty good and could (easily) win at the UAS competition. The team i am working with took second with paparazzi at the UAS this last year and we used IR. IR sensors also do have issues with fog and such but they are very nice because they do not suffer from drift

[dmgoedde]

Hey!! I am the developer of the AttoPilot, and I live RIGHT HERE in Gilbert next door to you. Atto is a thermopile autopilot similar to Paparazzi, and it is very accurate, and doesn't suffer from vibration sensitivity like an IMU does. Besides, I constantly cross-calibrate the thermopile raw data against barometric climb data and GPS turn rate... VERY accurate result.

Are you already set on the Paparazzi? No offense to them (they DO have a great thing), but I live right here next to you and could provide direct personal support. At the very least you could see a demo of the Atto in action.

Dean Goedde
AttoPilot LLC Manager

[dmgoedde]

The real problem with IR sensor is they are not calibratable in a real precise way. They have to be tuned by using your eye mostly rather a measuring device.I disagree completely. A 3-axis thermopile system with other data filtered in is 100% automatic with no calibration required at all.

[AntonK]

Dan,
You bring up a good point, that the calibration is not "accurate" but in actuallity a proceedure could be created to make it more accurate, it would just take time and some equipment. Because IMU based autopilots require integration the accuracy of the sensors is much more critical, and thus require either a higher cost IMU that is already calibrated, or some equipment to do the calibration yourself. Of course IR has its limitations(fog, clouds, outdoor only), but dont be fooled, you can fly over water, ice, in rain, and in mountainous areas(all these have been proven).

Eric, Luckily I dont work at the UA anymore, and actually left on a bad note, so I wouldnt mind helping the competition. We will see how the weekend shapes up as Ive been on the road for 6 weeks straight for work.

AJ

PS, well said Dean, and I agree he should see a demo of Atto, especially if their program has no desire to rewrite the code(as would be the advantage with paparazzi)

[dmgoedde]

For ability to modify code - definately go with a Paparazzi. If you just want to see what thermopiles can do before you get your Paparazzi up and running, I would be happy to give a demo.

I think FMA, AttoPilot LLC, and Paparazzi team must be the biggest 3 supporters and torch carriers for thermopiles!

[erichq2]

Hey!! I am the developer of the AttoPilot, and I live RIGHT HERE in Gilbert next door to you. Atto is a thermopile autopilot similar to Paparazzi, and it is very accurate, and doesn't suffer from vibration sensitivity like an IMU does. Besides, I constantly cross-calibrate the thermopile raw data against barometric climb data and GPS turn rate... VERY accurate result.

Are you already set on the Paparazzi? No offense to them (they DO have a great thing), but I live right here next to you and could provide direct personal support. At the very least you could see a demo of the Atto in action.

Dean Goedde
AttoPilot LLC Manager

Dean,

I would love to see a demo of AttoPilot. Would you mind PMing me your phone number and maybe we can get together sometime early this week? Actually, you can probably reach me easier at my email: chen.eric@asu.edu as that goes straight to my phone. I'm free after 12pm on Monday and Tuesday I'm busy, unless you want to come to ASU to demonstrate AttoPilot to our team meeting at 5pm.

No, we have not settled on Paparazzi yet but with a looming deadline of December 1st for an autonomous UAV using a prebuilt plane we do need to start making choices pretty soon so I hope we can work something out for at least a demo.

I did notice this little blurb on an interview you did though,

Telemetry jack to the XTend modem by Maxstream. This is currently being developed but is on track. Working now is the downlink to moving map software, but goal is 2-way com to take R/C control at long range, upload new waypoints, and issue commands such as RTL.

Is AttoPilot able to accept new waypoints or editable waypoints midflight yet? That's one of the requirements of the AUVSI UAS competition and if AttoPilot can't do that then we couldn't use it.

I really like the idea of being next door to the developer, though.

[Gary Mortimer]

Blimey thank goodness setting up UAV's is so simple otherwise I might be worried about a Dec 1 deadline!!

Students eh, the same the world over!

[dmgoedde]

Dean,

I would love to see a demo of AttoPilot. Would you mind PMing me your phone number and maybe we can get together sometime early this week? Actually, you can probably reach me easier at my email: chen.eric@asu.edu as that goes straight to my phone. I'm free after 12pm on Monday and Tuesday I'm busy, unless you want to come to ASU to demonstrate AttoPilot to our team meeting at 5pm.

No, we have not settled on Paparazzi yet but with a looming deadline of December 1st for an autonomous UAV using a prebuilt plane we do need to start making choices pretty soon so I hope we can work something out for at least a demo.

I did notice this little blurb on an interview you did though,



Is AttoPilot able to accept new waypoints or editable waypoints midflight yet? That's one of the requirements of the AUVSI UAS competition and if AttoPilot can't do that then we couldn't use it.

I really like the idea of being next door to the developer, though.
Early this coming week is not good.... but the week after that probably. Yes, you will be able to upload new waypoints and do all sorts of commands during flight.

Keep in mind Atto is closed source, and Paparazzi is open. With help on these forums I believe you could get a Paparazzi up and running before your deadline.

I will PM you. You are catching me at the end of a 2 month development string with no flying going on... BUT my assembler AltitudeAP lives in Mesa and has some Atto-equipped planes flyable right now.

[Mecha]

I think this competition is going to have some heat this June. I am a Mechanical Engineer student at Florida International University and for my senior design project I will be working on a uav to enter the AUVSI UAS
FIU has never been involved in any uav competitions and we currently only have a minor in AE that just began this year. Hopefully this project will enhance our AE program.

So far the project is entirely funded by me and my other team mate. I am relying heavily in the fact that I have been flying planes and helicopters for many years and I have a lot of RC equipment already at hand. I am working with the paparazzi autopilot for the same reason as you, that is, price.
I am waiting patiently for dmgoedde to release his Atto RTL so I can give it a shot.

dmgoedde I wish I lived closer to Arizona so that I could also see the Atto in action. Does any of your test pilots live in Florida? or can you guys provide me with some video of the system and the software?

I do not mean to hijack your thread and once I begin construction I will open a new thread documenting the project progress.

[ALtitudeap]

Mecha, Where are you at in florida? I will be traveling to Florida on the 26th of this month. If you have a plane, i can bring my Attopilot and make you plane fly a demo.

[dmgoedde]

dmgoedde I wish I lived closer to Arizona so that I could also see the Atto in action. Does any of your test pilots live in Florida? or can you guys provide me with some video of the system and the software?

I do not mean to hijack your thread and once I begin construction I will open a new thread documenting the project progress.I have 2 testers in Florida and they are active participants. The towns are Ponte Vedra Beach, and Miami. PM me for more details.

[Mecha]

Guys many thanks for the quick replays..

ALtitudeap, I am in Dade County, Miami. But I will be willing to travel during the weekend within reason (200-250 miles) to see a demo.


mgoedde PM sent.

[erichq2]

Well, we have ordered Tiny 2.11 and it should arrive here either Friday or very early next week from what I can tell. A few of us have been playing around with Paparazzi, trying to figure out how to install Linux, etc. We have tons of help from friends of friends who know a few things about rc aircraft, so we're working on getting a good weight estimate so our airframe team can start writing code and design a fuselage.

We'll probably go with a high aspect wing ratio type sailplane to start out with. Though further iterations I'd like to create something more... exotic.

[AntonK]

you should buy a Multiplex Cularis to start with. It flies really nice, and would be a good paparazzi starter while you get your stuff going.

AJ

[erichq2]

We already have a plane that we can use from a guy on the team. I'm not sure what plane it is though. He's never flown it before but I think we found ourselves a safety pilot or somebody to teach us how to fly.

I'm also trying to find out what IMU our faculty mentor has for us so that I can talk to you about IMU integration.

[Tuner]

Eric Im confused here?
So is this part of http://www.uofaarc.org/

The reason I ask is I just started at UA and I had planned on joining this club.

Scott

[erichq2]

Eric Im confused here?
So is this part of http://www.uofaarc.org/

The reason I ask is I just started at UA and I had planned on joining this club.

Scott

I'm afraid not. This is a project at ASU.

[AntonK]

Tuner,
you are at the UofA now? ARC doesnt use paparazzi, they use Piccolo, have for years, I used to work with the founder of the club.

AJ

[Tuner]

I'm a idiot, yah UA ASU its all the same to me I just want army of robot planes :)
Duh!
Hopefully see you guys in June then!


I did not know that for sure I suspected from their sponsor list? I always assumed they used Paparazzi because I was talking with a Jeremy whom I assumed attended UA.

Well Anton they might start using Paparazzi If I have anything to say about it. I am anticipating that if I supply all my own hardware that they will have no problem letting me use it on my own bird. I currently have an Atiar Aerospace IMU/INU that I am dying to test out with the classix and maybe the tiny. I am looking forward to learning about the differences between a high end IMU and a well tuned IR setup. If they don't I will just work on it separately. I want any of my work at school to translate in to an eventual business and you don't make money licensing other products and its hard to be unique when you cant modify the foundation of your product.

Infact since I have your attention I want to buy a Tiny and Classix the tiny is available but the classix is more elusive. I have talked with a local company in Phoenix that does assembly and is a decent deal for small runs but I am hoping to maybe buy a used or built/tested board as I don't really have the time or patients.

I am going to attend a meeting for people interested in ARC tonight. I have emailed these guys without response so many times so this should work.

[erichq2]

The IMU we have available is a MicroStrain 3DM-GX1 (http://microstrain.com/3dm-gx1.aspx). Sound familiar?

[AntonK]

Tuner,
I still live in Tucson(Jeremy has since moved back to california), we should meet up sometime. I could go to an ARC meeting or something.

erich,
Ive actually flown with that IMU before on a quadrotor, its a great little IMU. The nice thing about it is that it does all the filtering for you, and outputs roll pitch and yaw. You have little to do as far as integration besides setting up the device drivers and polling the data. If its TTL serial you may run into a small snag as the 2 that are on the Arm7 are currently being used. However if its SPI or I2C then you are in business.

AJ

PS. Tuner you dont want a classix, trust me.