I am looking to build my first quad copter as a fully autonomous GPS system. I was looking at the pixhawk systems and am liking that but would appreciate any advice as far as that and the whole project goes.... I need to buy everything from the frame up and need reccomendations. This is my first quad-copter build and willl be used mostly for surveillance but I also need to be able to program it to fly to a pre determined destination, land, and return home either on command or after a couple minute timer. I am not a complete rookie however as I have been an r/c hobbyist for years but this will be first fully autonomous set up.... Thank you in advance for any offers of help!
Help building autonomous GPS quad
- Thread starter bda714
- Start date
Bartman
Welcome to MultiRotorForums.com!!
you're about to embark on a long and sometimes difficult project. start out by building something simple so you can learn what the different components do and then work on adding capabilities as you develop the knowledge necessary to make things work. it would take us hours and hours in hundreds of posts to teach you everything that you need to know.
check out the Group Build thread for an idea of how to get started.
check out the Group Build thread for an idea of how to get started.
RaptorBird Robotics
Member
Hello, we have a drone development platform which allows for programming takeoff, autonomous flight, and land sequences. It's written in Python and you would use our API which makes it intuitive to do exactly what you said. Our hardware and software include the necessary components to execute your mission as easily as possible.
It also works on any drone form.
Please take a look.
http://www.kickstarter.com/projects/raptorbirdrobotics/ravn-easy-drone-development
It also works on any drone form.
Please take a look.
http://www.kickstarter.com/projects/raptorbirdrobotics/ravn-easy-drone-development
Av8Chuck
Member
Not sure I understand what this is? It works with Pixhawk? So it just manages the mission control and not a flight controller?
RaptorBird Robotics
Member
It's an add on that makes it easy to program missions, that way you can include other code alongside the missions. Like adding computer vision code using existing frameworks such as SimpleCV/OpenCV to monitor the drone's surroundings and make decisions based on environmental input.
It basically turns your drone into a flying robot capable of intelligent decision making.
Not only that, but it's very easy to get started.
It works with the Pixhawk flight controller. So yes your drone would need a Pixhawk so that basic flight control is taken care of.
By having a Pixhawk configured for your drone type the code you write for the RAVN doesn't change, which makes development extremely portable.
It basically turns your drone into a flying robot capable of intelligent decision making.
Not only that, but it's very easy to get started.
It works with the Pixhawk flight controller. So yes your drone would need a Pixhawk so that basic flight control is taken care of.
By having a Pixhawk configured for your drone type the code you write for the RAVN doesn't change, which makes development extremely portable.
Gary Seven
Rocketman
I'm sorry. Color me jaded...color me pessimistic. Your post (in my eyes at least) borders on SPAM. What's more, you have about a snowball's chance in hell of that thing (essentially an autonomous robot) being permitted to fly in any airspace controlled by the FAA.
I stand ready to be corrected on either of my points above.
RaptorBird Robotics
Member
Gary,
We are going to require that the user has a remote connected to the multirotor to switch on the mode that lets our software run. That way there's always a human present while the drone is running the sequence. This allows the drone to commit to the autonomous task while still having a human ensure that the task runs smoothly. The person can always override our system and safely land or fly the drone to a safer area to land.
I can see your concerm, but we're a small startup comprised of two college students. We're just trying to build a following around our product which we believe can help many people build drone systems to make their lives easier.
But don't just take it from me, you should please check out the Kickstarter and let us know what you think.
Thanks!
We are going to require that the user has a remote connected to the multirotor to switch on the mode that lets our software run. That way there's always a human present while the drone is running the sequence. This allows the drone to commit to the autonomous task while still having a human ensure that the task runs smoothly. The person can always override our system and safely land or fly the drone to a safer area to land.
I can see your concerm, but we're a small startup comprised of two college students. We're just trying to build a following around our product which we believe can help many people build drone systems to make their lives easier.
But don't just take it from me, you should please check out the Kickstarter and let us know what you think.
Thanks!
Bartman
Welcome to MultiRotorForums.com!!
@RaptorBird Robotics; my only issue as the site Admin with how you're going about things is that every post of yours is a plug for your kickstarter campaign. i hope you're successful but you won't make a lot of friends if all you are going to do is plug your campaign every chance you get without any other participation at the site.
i'd suggest, if you haven't done it already, that you start a thread titled appropriately for your project and that in that thread you list various accomplishments that you've been able to achieve with stuff from your project. use it as a place to also answer specific questions from our membership and you'll begin to win over people who will consider investing in your campaign.
if you continue to plug your project wherever possible then we'll have to figure out a way to deal with that.
good luck with it.
i'd suggest, if you haven't done it already, that you start a thread titled appropriately for your project and that in that thread you list various accomplishments that you've been able to achieve with stuff from your project. use it as a place to also answer specific questions from our membership and you'll begin to win over people who will consider investing in your campaign.
if you continue to plug your project wherever possible then we'll have to figure out a way to deal with that.
good luck with it.
RaptorBird Robotics
Member
Av8Chuck
Member
I don't think its an issue of whether Bart approves, probably more of an explanation of why your being met with some skepticism. Nothing wrong with being excited about what your developing but try to explain to people why you think it will benefit them. And keep in mind that as a new member you lack credibility, which again doesn't mean that your not onto something good.
Also, and I'm not being a smart a$$, but I'm not totally sure what this is or why you designed it, and this is after I looked at your kickstarter. So you might want to give some thought to your messaging because whatever you think your saying might not be coming across effectively.
I think you should start a thread for this but keep in mind that it might be a good idea to provide information about yourself and your company so that we can get a sense of who you are what your doing, tell us a little of why you think what your doing is necessary and ask people what they think about it. There are people on this forum that have been building MR's from before there were MR's, which can make them a valuable resource for you to learn from and most don't mind sharing their experience but if you reduce the debate to a sales proposition then your reducing the likelihood that you'll have the opportunity to benefit from their experience.
Also, and I'm not being a smart a$$, but I'm not totally sure what this is or why you designed it, and this is after I looked at your kickstarter. So you might want to give some thought to your messaging because whatever you think your saying might not be coming across effectively.
I think you should start a thread for this but keep in mind that it might be a good idea to provide information about yourself and your company so that we can get a sense of who you are what your doing, tell us a little of why you think what your doing is necessary and ask people what they think about it. There are people on this forum that have been building MR's from before there were MR's, which can make them a valuable resource for you to learn from and most don't mind sharing their experience but if you reduce the debate to a sales proposition then your reducing the likelihood that you'll have the opportunity to benefit from their experience.
RaptorBird Robotics
Member
@Av8Chuck
Thank you for the suggestions. If I send you a private message implementing the advice you just said would you be willing to look it over to see if I'm headed in the right direction? That way I can get it right before posting publicly.
If you have a moment to look it over that would be very helpful.
Thank you!
Thank you for the suggestions. If I send you a private message implementing the advice you just said would you be willing to look it over to see if I'm headed in the right direction? That way I can get it right before posting publicly.
If you have a moment to look it over that would be very helpful.
Thank you!
Av8Chuck
Member
Sure, but don't over think it. When your starting a business or launching a product, going through this very process is important, it helps you develop your message, along with feedback for dealing with criticism, as well as think of things you might not have innitially thought of.
So just start your thread, be honest about "Why" your doing this, explain a little about how your doing it, you don't have to give away anything proprietary, I (for example) only want to know how it works with the Pixhawk? Does it make setting up the Pixhawk easier? Does it make it easier to create a mission plan than what's already available? Who is the target audience for it? My guess from what I understand is a programer or developer who wants to add value to the ability of the FC.
So just start your thread, be honest about "Why" your doing this, explain a little about how your doing it, you don't have to give away anything proprietary, I (for example) only want to know how it works with the Pixhawk? Does it make setting up the Pixhawk easier? Does it make it easier to create a mission plan than what's already available? Who is the target audience for it? My guess from what I understand is a programer or developer who wants to add value to the ability of the FC.
RaptorBird Robotics
Member
JoeBob
Elevation via Flatulation
The Pixhawk flies autonomous missions very well. But there are limitations:
1- GPS position accuracy is 3 to 4 meters from your waypoint, not 3 to 4 feet like I was hoping. Don't plan on zooming through narrow gaps. (Even though you can set your Waypoint radius to a small value, that small waypoint can be a few meters off.)
2- Altitude is based on air pressure and GPS. You can tell the Pixhawk how to balance input between the baro and the GPS according to your application. But it is still not accurate enough for low level (1 meter) terrain following. I'm working with sonar sensors and investigating lidar for better terrain following.
3- Mission Planner is the free application that supports Pixhawk, and it is very good. There are about 1000 different inputs and toggling any of them to a wrong setting can crash you. My only complaint about Mission Planner and Pixhawk is the absence of a current primer. The program is open source and constantly being upgraded, so you have to understand that having a volunteer update a 'latest' Wiki would be hard. Check the dates on any material you are reading so that you won't waste time following an obsolete info trail that will live forever on the internet.
4- There are 2 support forums that you'll have to spend a few days reading to educate yourself. Sometimes your inquiries will get a response in hours, sometimes in weeks; and sometimes you'll get bad advice.
5- 'Spline Waypoints' and 'Point of Interest' are great and produce a much smoother flight/video than most folks can get flying manually. I'll fire up the 'Auto' flight mode and then focus on my camera tilt.
6- 'Return to Launch' is the flight mode for recovery. You can set RTL as a flight mode for emergencies or program RTL as the last feature of a mission. RTL comes straight back to your launch point, so if you have obstacles on your route, set an appropriate height for RTL.
7- 'Land' is my favorite flight mode...
8- As Bartman pointed out, building and flying MRs is complicated. A Pixhawk would be a good second controller, after you've learned the basics.
1- GPS position accuracy is 3 to 4 meters from your waypoint, not 3 to 4 feet like I was hoping. Don't plan on zooming through narrow gaps. (Even though you can set your Waypoint radius to a small value, that small waypoint can be a few meters off.)
2- Altitude is based on air pressure and GPS. You can tell the Pixhawk how to balance input between the baro and the GPS according to your application. But it is still not accurate enough for low level (1 meter) terrain following. I'm working with sonar sensors and investigating lidar for better terrain following.
3- Mission Planner is the free application that supports Pixhawk, and it is very good. There are about 1000 different inputs and toggling any of them to a wrong setting can crash you. My only complaint about Mission Planner and Pixhawk is the absence of a current primer. The program is open source and constantly being upgraded, so you have to understand that having a volunteer update a 'latest' Wiki would be hard. Check the dates on any material you are reading so that you won't waste time following an obsolete info trail that will live forever on the internet.
4- There are 2 support forums that you'll have to spend a few days reading to educate yourself. Sometimes your inquiries will get a response in hours, sometimes in weeks; and sometimes you'll get bad advice.
5- 'Spline Waypoints' and 'Point of Interest' are great and produce a much smoother flight/video than most folks can get flying manually. I'll fire up the 'Auto' flight mode and then focus on my camera tilt.
6- 'Return to Launch' is the flight mode for recovery. You can set RTL as a flight mode for emergencies or program RTL as the last feature of a mission. RTL comes straight back to your launch point, so if you have obstacles on your route, set an appropriate height for RTL.
7- 'Land' is my favorite flight mode...
8- As Bartman pointed out, building and flying MRs is complicated. A Pixhawk would be a good second controller, after you've learned the basics.
DesignTypeGeek
Quad-copter Newbie
Too bad the Kickstarter got cancelled.
LMAO. . . @Gary Seven - 1984 called and they said the flux capacitor is fixed now. You are cleared for a return trip in your delorian. . .
Ref: http://ardupilot.com/ https://wiki.paparazziuav.org/
You stand by to be corrected on your points huh? Well while you are standing by waiting to be corrected life is passing you by. Obviously you don't keep up to date by reading anything except forums trolling for who knows what... *cough* ahhm @Gary Seven = #toCatchAPredator *cough* excuse me, but ya try reading something that can be qualified as non-fiction some time. AFAIK, at the time of my post there was no legislation passed on the issue, now however there is, and it doesn't say anything about making this illegal. The FAA has brought many regulations, many of which are very vague, but even at the most basic level, you might be able to understand some of it. There is altitude regulations, airspace, line of sight, but no melting snowballs as of yet.
Now thank you to the few that actually had some facts or info to attribute, I really appreciate it. I have since finished a build using an ardupilot w/ GPS that I picked up from 3DRobotics. Very cool front end that allows for everything that I was looking for... I brought out an old Futaba 7 chan radio and got a iFly kit which has been awesome. Haven't had to use the radio control though even for a second (= perfect for a drone. Really if I want to control something I will take out a standard heli or plane. The drone is just that, a drone. Sorry to see my post has transformed into a shameless plug for a failed kickstarter and 1 D-Bag that is juggling snowballs to kill time while waiting for someone to educate him. But to everyone not lame. thanks for the info and the community here is a great wealth of knowledge. If only we all took advantage of it for the right reasons, the possibilities would be endless.
@Gary Seven - Classic response, You made my day! thank you. You mad bro? Come on bro, you mad?. . . don't you know that is the exact response anyone who is trolling is looking for, Don't worry about being clever or witty, Nah, just throw a bunch of f bombs out and that should get your point across. You welcomed my comment, sir, by stating you were "standing by to be corrected". Now you've been corrected and shown that trying to sound smart in a public venue probably isn't a good idea for a person at your iq level. . . even worse your e.i. level... Next time take a deep breath sir, and think a minute (but dont hurt your self) and only then reply. If you can do that, and get to a level that i can at least not have to strain my neck to badly looking down for you, then I'll consider debating you on any subject, until then sir, you are just a form of amusement for folks with any sense of reason or intelligence. Try hijacking someone else's post who will just take a guy like your's word for it... P.s. - nice pic, #D-bag-selfies #to-catch-a-predator.