My robotics project is an autonomous go-kart. GPS waypoints will be input and the robot will use an onboard GPS reciever and a PING))) sensor with servo bracket to navigate to it. Type what you think!
Sounds dangerous. But then again it might be fun to see it take out neighbors mailboxes. Ha!. Ahh..no seriously, I'd make sure to test it on a secluded track or road. Sounds like fun though. Good luck.
Dave
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ "THE ONLY TRUE WISDOM IS IN KNOWING YOU KNOW NOTHING." - SOCRATES
Just a question - have you tested the accuracy of the GPS? If the accuracy is only good to 20 feet, how will you compensate for navigating a 10-foot wide road or pulling up to a mailbox that requires a much greater precision? Or is the GPS accuracy much better and improved on the latest models and mathematical algorithms?
It sounds like a fun and expensive project. No such project is a small undertaking, you will for sure need to do your home work!
I would suggest building it up in small sub projects. Use a little wheeled robot first. Build the navigation system and test it. Then build the sensor array of PING))) and other sensors. Mount it to the go-cart and adjust the sensors based on you driving. Then add in the system controlling the motors and such. Oh yeah, not to mention the steering and braking. WOW! A lot to do.
Be careful.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Timothy D. Swieter
www.brilldea.com·- check out the uOLED-IOC, an I/O expansion for the uOLED-96-PROP www.tdswieter.com
One little spark of imagination is all it takes for an idea to explode
R/C cars or trucks are a good starting point for a project like this. Build the navigation system and hack it into the steering while you keep manual control of the throttle. This is a great way to test stuff out without risking your expensive GPS electronics getting away from you and crashing into something!
that's a Project that I had plans to do.... in the distant future. I was thinking more lawn mower than go cart though, but in my opinion no robot is too big. I think Sensor Range is going to be your biggest hurdle.
you did mention that you wanted discussion not feasibility so let's discuss this, and get you moving in the direction you want to go. Can you give more information? What size are you talking about, is this going to be gas powered or electric, what type of steering are planning on using. What speeds are you hoping for.
Honestly I wouldn't bother with scale model testing unless it's something like, Differential steering vs rack & pinion. in my experience the issues at the boebot scale aren't proportional in a larger scale. there are also a whole new set of unknowns. If starting with a boebot is to learn Pbasic or electronics, the I would definitely start there. and I second Timothy D. Swieter on breaking it down to smaller project get it working and then go to the next.
I can't commit to a lot of time or revisions but when you get to the point, I would be happy to produce some CAD drawings a 3D model in sketchup. for you.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"A complex design is the sign of an inferior designer." - Jamie Hyneman, Myth Buster
Dgswaner- Thanks for offering to help! Here is my plan so far:
A Go-kart (gasoline-powered) will be outfitted with a PING))) Ultrasonic sensor(on a servo bracket) and a GPS receiver (more sensors will be added later on). A cam will be mounted to a servo to control the speed and brake(thanks PJ!).
I believe that one of my largest problems is steering, here is my idea:
I will remove the front two go-kart wheels. I will then attach a bike frame and front wheel. A bar will be attached to the servo. Two wires would then be attached to either side of the bar and the handles on the bike. The servo will turn the handlebars in the necessary direction.
A couple of things that you might want to consider, depending on the size and speed of your kart (karts come in a huge range of capabilities) A totally separate kill switch (remote-controlled) might be nice (especially if there's a glitch and it heads for the neighbor's new Lexus (or worse, the neighbor!)
Also, the steering loads on a kart big enough to hold a person, even a child, can be substantial. I would NOT expect the standard hobby servo to do the job. The yoke-and-cable arrangement sounds feasible, but you might want to think about running the cable to a small drum and have a motor rotate the drum. Springs to pre-load the steering to encourage self-neutral might be good as well. I guess that the steering issues are a good example of the sort of thing that wouldn't scale up well...
Curious, I think it will be advantageous to keep the go cart operable by a human, sitting in it, imagine the scenario of being in a field and a processor dies, fries or what ever and you'll have to push and or steer the go cart back. If the Mythbusters can do it with a full blown car, I think it's totally doable with a go cart. I would actually recommend watching the episode where they RC a police car, they show how to do exactly what your after.
Steering? I find that a Screw drive makes a very good, very stable way to move something, even a round steering wheel. our you could use a stepper motor with a reduction gear. they do sell large servos capable of doing the job.
on the throttle, a Hobie servo should be able to handle the throttle, if you controlled it at the motor, the throttle arm is quite easy to move on most small engines.
it would be expensive but you could also go hydrolic and have it handle every thing.
let us know what your plans are
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"A complex design is the sign of an inferior designer." - Jamie Hyneman, Myth Buster
curious roboticist said...
I believe that one of my largest problems is steering, here is my idea:
I will remove the front two go-kart wheels. I will then attach a bike frame and front wheel. A bar will be attached to the servo. Two wires would then be attached to either side of the bar and the handles on the bike. The servo will turn the handlebars in the necessary direction.
I don't see any advantage to removing the go-kart front end and replacing with single bike wheel. Any steering method you chose could just as easily be accomplished at the go-kart steering wheel as a bike wheel.
Dgswaner's suggestion of a screw gear is a good possibility. I would suggest you consider two things, one is the amount of torque you require. Placing a larger (10-12 inch diamemeter gear on the underside of the steering wheel and driving it with a small screw drive or pinnion drive gear MIGHT give enough torque. The down side of such a high gear reduction is the·response time to make course adjustments might happen to slowly. That is, you can turn the wheel right to miss the neighbors lexis, but it might take three or four seconds to make the·wheel complete a turn.(understeer) A go-kart could probably travel quite a distance in three seconds. This might lead to crashes·or slow course corrections. In summary get ALL the speed and torque you can possibly afford!
Have any pictures of the go-kart you are planning to convert? specifically the existing steering linkage, so we can see what we have to work with. I remember the go-kart I had as a kid, and adapting to a single front wheel would have been next to impossible.
The rods that attach the steering knuckles to the base of the shaft the connects to the steering wheel are not a nice ball joint type connection, they are simply just a rod stuck through a hole. Doesn't make for easy smooth steering. I would suggest upgrading the linkage to better joints if they are not already. Maybe change the length of the arm that the rods attach to to make the steering response more linear (moving the steering wheel moves the wheels by an identical angle each time, regardless of starting position)
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Brian
uController.com - home of SpinStudio - the modular Development system for the Propeller
PropNIC - Add ethernet ability to your Propeller! PropJoy - Plug in a joystick and play some games!
You could use bicycle type sprockets on the steering shaft and a smaller sprocket on a servo type motor. This should handle the torque required to turn the steering wheel.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ David
There are 10 types of people in this world,... Those that understand binary numbers, and those that don't!!!
Hey. Sorry I haven't been on for a while, I have been really occupied. Anyway, I found some linear actuators on servocity.com. How could I control them since they are motors, not servos?
If you are looking for lots of torque, you might want to investigate this route. I just finished building a servo to steer a Power Wheels Safari Jeep out of a windshield wiper motor (rear window of a minivan) and a POT.
The motor driver/controller is a Pololu SMC04B , which has a mode that directly positions a motor using a pot and controlled by servo-type pulses or serial input. right now I'm using servo pulses.
I once saw a guy use this model motor to swing a baseball bat, so I was aware of what it could do.
I'll admit I was a little nervous the POT would get destroyed the first time I energized things.
Because of the direct drive I used I wouldn't get second chances.
I epoxied the Pot into a hole I drilled in the back of the shaft because I couldn't find a way to use a set screw (the drive gear is right there).
If I were to do it again I might de-couple things until I had tested,
but everything worked perfectly the first time.
I would suggest adding bump sesors to the front and read as well as IR. The idea being if PING))) experinces some weird problems for some reason, IR could be used if it's dark enough, if all else fails the bump sensor could at least stop it from driving after impact, if they both fail, to minimize damage. Both a remote kill switch decoupled from the processor and a very visible kill switch on the go-kart are a good idea, both the MythBusters and BattleBots agree.
Comments
Dave
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"THE ONLY TRUE WISDOM IS IN KNOWING YOU KNOW NOTHING." - SOCRATES
humanoido
I would suggest building it up in small sub projects. Use a little wheeled robot first. Build the navigation system and test it. Then build the sensor array of PING))) and other sensors. Mount it to the go-cart and adjust the sensors based on you driving. Then add in the system controlling the motors and such. Oh yeah, not to mention the steering and braking. WOW! A lot to do.
Be careful.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Timothy D. Swieter
www.brilldea.com·- check out the uOLED-IOC, an I/O expansion for the uOLED-96-PROP
www.tdswieter.com
One little spark of imagination is all it takes for an idea to explode
Mike
you did mention that you wanted discussion not feasibility so let's discuss this, and get you moving in the direction you want to go. Can you give more information? What size are you talking about, is this going to be gas powered or electric, what type of steering are planning on using. What speeds are you hoping for.
Honestly I wouldn't bother with scale model testing unless it's something like, Differential steering vs rack & pinion. in my experience the issues at the boebot scale aren't proportional in a larger scale. there are also a whole new set of unknowns. If starting with a boebot is to learn Pbasic or electronics, the I would definitely start there. and I second Timothy D. Swieter on breaking it down to smaller project get it working and then go to the next.
I can't commit to a lot of time or revisions but when you get to the point, I would be happy to produce some CAD drawings a 3D model in sketchup. for you.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"A complex design is the sign of an inferior designer." - Jamie Hyneman, Myth Buster
DGSwaner
A Go-kart (gasoline-powered) will be outfitted with a PING))) Ultrasonic sensor(on a servo bracket) and a GPS receiver (more sensors will be added later on). A cam will be mounted to a servo to control the speed and brake(thanks PJ!).
I believe that one of my largest problems is steering, here is my idea:
I will remove the front two go-kart wheels. I will then attach a bike frame and front wheel. A bar will be attached to the servo. Two wires would then be attached to either side of the bar and the handles on the bike. The servo will turn the handlebars in the necessary direction.
Grievousfish-Yeah, the Quadrover is similar.
A couple of things that you might want to consider, depending on the size and speed of your kart (karts come in a huge range of capabilities) A totally separate kill switch (remote-controlled) might be nice (especially if there's a glitch and it heads for the neighbor's new Lexus (or worse, the neighbor!)
Also, the steering loads on a kart big enough to hold a person, even a child, can be substantial. I would NOT expect the standard hobby servo to do the job. The yoke-and-cable arrangement sounds feasible, but you might want to think about running the cable to a small drum and have a motor rotate the drum. Springs to pre-load the steering to encourage self-neutral might be good as well. I guess that the steering issues are a good example of the sort of thing that wouldn't scale up well...
Have fun,
Terry
Steering? I find that a Screw drive makes a very good, very stable way to move something, even a round steering wheel. our you could use a stepper motor with a reduction gear. they do sell large servos capable of doing the job.
on the throttle, a Hobie servo should be able to handle the throttle, if you controlled it at the motor, the throttle arm is quite easy to move on most small engines.
it would be expensive but you could also go hydrolic and have it handle every thing.
let us know what your plans are
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"A complex design is the sign of an inferior designer." - Jamie Hyneman, Myth Buster
DGSwaner
Dgswaner's suggestion of a screw gear is a good possibility. I would suggest you consider two things, one is the amount of torque you require. Placing a larger (10-12 inch diamemeter gear on the underside of the steering wheel and driving it with a small screw drive or pinnion drive gear MIGHT give enough torque. The down side of such a high gear reduction is the·response time to make course adjustments might happen to slowly. That is, you can turn the wheel right to miss the neighbors lexis, but it might take three or four seconds to make the·wheel complete a turn.(understeer) A go-kart could probably travel quite a distance in three seconds. This might lead to crashes·or slow course corrections. In summary get ALL the speed and torque you can possibly afford!
Ted
Dgswaner- I have some old scavenged stepper motors at home so I might use those. Let me look and I will get back to you.
http://www.youtube.com/watch?v=ulpnT5_oQww
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·"If you build it, they will come."
Big kudos for starting a great discussion.
First and foremost, as Terry mentioned, how do you stop it when it goes wrong!
Second, be a little careful how fast you plan on going. 3m is a fair distance, but how fast does your vehicle react when something is detected?
Third, I liked the screw drive (DGS), but why not access the bar linking the wheels directly?
The rods that attach the steering knuckles to the base of the shaft the connects to the steering wheel are not a nice ball joint type connection, they are simply just a rod stuck through a hole. Doesn't make for easy smooth steering. I would suggest upgrading the linkage to better joints if they are not already. Maybe change the length of the arm that the rods attach to to make the steering response more linear (moving the steering wheel moves the wheels by an identical angle each time, regardless of starting position)
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Brian
uController.com - home of SpinStudio - the modular Development system for the Propeller
PropNIC - Add ethernet ability to your Propeller! PropJoy - Plug in a joystick and play some games!
SD card Adapter - mass storage for the masses Audio/Video adapter add composite video and sound to your Proto Board
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
David
There are 10 types of people in this world,...
Those that understand binary numbers, and those that don't!!!
The motor driver/controller is a Pololu SMC04B , which has a mode that directly positions a motor using a pot and controlled by servo-type pulses or serial input. right now I'm using servo pulses.
Enough torque to break your arm.
Post Edited (Larry) : 5/6/2008 7:32:56 AM GMT
I once saw a guy use this model motor to swing a baseball bat, so I was aware of what it could do.
I'll admit I was a little nervous the POT would get destroyed the first time I energized things.
Because of the direct drive I used I wouldn't get second chances.
I epoxied the Pot into a hole I drilled in the back of the shaft because I couldn't find a way to use a set screw (the drive gear is right there).
If I were to do it again I might de-couple things until I had tested,
but everything worked perfectly the first time.
Post Edited (Larry) : 5/7/2008 4:40:00 PM GMT