Shop OBEX P1 Docs P2 Docs Learn Events
Robot Lawn Mower — Parallax Forums

Robot Lawn Mower

chunnchunn Posts: 6
edited 2007-08-18 11:22 in Robotics
I sure a lot of people have done this already but part of the fun is doing it myself. I'm currently planning on using the BASIC stamp microprocessor with the PINK module for internet access to the robot and the GPS module for location. I plan on using two linear actuators to control the two arms of my TORO zero turn mower.

I would appreciate any advise before I begin with this project. Here are some of the things planned.

Most of the settings for the unit will be initially set via the web interface using the PINK module.

My goal is to take the mower out in my yard ( a total of 9 acres that I'm mowing )

maps.google.com/maps?f=q&hl=en&geocode=&q=18745+Canoebrook+Ln,+Toney,+AL+35773&sll=37.0625,-95.677068&sspn=46.36116,82.265625&ie=UTF8&ll=34.82617,-86.814694&spn=0.011802,0.020084&t=k&z=16&iwloc=addr&om=1

and set four reference points that will create a rectangle that will be the bounds for the mower to mow inside of. Once the four points are set you will be able to visit the web interface for the robot mower and get a link to google maps that will show the rectangle box.

mow.applehat.com

The link above is a working interface that currently just displays the last area registered to mow.

One concern is using GPS to keep the robot inside the rectangle. Another concern is the actuator and it's control. I plan on using one that has a potentiometer so that I know where the control arm is at.

Also I'm concern about talking to the GPS and PINK unit at the same time. I'm assuming that they both listen to the data on the serial out and only respond if they recognize it. The GPS is fixed at 4800 baud but I think the PINK module can be configured for 4800 baud. I'm assuming that they both need to be at the same rate.

Thanks for any help on this project.

Post Edited (chunn) : 8/15/2007 4:03:30 AM GMT
282 x 212 - 23K
282 x 212 - 28K

Comments

  • chunnchunn Posts: 6
    edited 2007-08-15 04:05
    In the second picture you see the control arm that connects to the Hydrostatic drive. This is where I would like to attach an actuator. That would allow me to connect it to the metal frame of the mower and would only need 1.5 inch stroke on the actuator. The only disadvantage is that it would need around 30-40 lbs of force to move the arm versus about 10-15 lbs of force if I use the arm control level on top of the mower.

    One problem to solve is to make it easy to disengage the actuators so that I can use the mower as intended.
  • chunnchunn Posts: 6
    edited 2007-08-16 21:11
    Nobody can help here?
  • Mike GreenMike Green Posts: 23,101
    edited 2007-08-16 21:56
    The GPS unit and the PINK can use different Bauds or they can be the same. You can't talk (or listen) to both at the same time. It's strictly one at a time and one direction at a time.

    I don't think that GPS will give you accurate enough locations for mowing. You will probably need markers of some sort at the periphery, maybe IR beacons or PING looking for specific posts or something like that. GPS may end up being of little use.
  • Chris SavageChris Savage Parallax Engineering Posts: 14,406
    edited 2007-08-17 03:18
    Cross Posts in the BASIC Stamp forum have been removed. Please bear in mind that if your post is clear most forum members will reply if they feel they have something to offer.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Chris Savage
    Parallax Tech Support
  • bobledouxbobledoux Posts: 187
    edited 2007-08-17 14:32
    I'm reluctant to reply because I'm not certain you understand the ramifications of your project.

    You have a multi-horsepower lawn mower, a very dangerous machine, operating without direct human control. What happens if it encounters an unexpected object, like a small child? Does it just run over the object? Your brief question doesn't give me confidence you are considering this fundamental issue.

    There is a good reason robot mowers use a random movement within a limited area of operation. That's because there is no simple way to provide path control required by a mower. Straight GPS is too slow and inaccurate to provide constant path control. It may be simple to navigate a robot from point A to point B, but a mower also must monitor the path between the two points to achieve the mowing pattern. Achieving that goal is not a trivial pursuit.

    Here is one approach: http://www.landscapemanagement.net/landscape/article/articleDetail.jsp?id=169360
  • Jim RicheyJim Richey Posts: 82
    edited 2007-08-17 16:17
    Hey,Chunn

    Glad to see you're thinking.That's a challenging project to accomplish.
    I thought you might like to see a couple shots of my remote control lawnmower.
    It's been in full use for 3 years,I don't know how I ever managed to cut the
    lawn with out it as my yard is very steep and dangerous to even walk on.
    This machine,even with fully functional controls,has been upside down several times,
    on it's side,it's even tumbled end over end [noparse][[/noparse]that's the reason for the wheelie and roll bars].
    It would seem dangerous beyond imagination to allow something like this to roam around
    with nothing controlling it except for a puny silicon brain.Lawnmowers especially are
    capable of doing damage and destruction at a moments notice.

    Be Careful!! Jim Richey

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Thanks, Parallax!
    2048 x 1536 - 1M
    2048 x 1536 - 1M
  • bobledouxbobledoux Posts: 187
    edited 2007-08-17 16:55
    Jim, Why two big batteries?
  • chunnchunn Posts: 6
    edited 2007-08-17 16:56
    Mike Green said...
    The GPS unit and the PINK can use different Bauds or they can be the same. You can't talk (or listen) to both at the same time. It's strictly one at a time and one direction at a time.

    I don't think that GPS will give you accurate enough locations for mowing. You will probably need markers of some sort at the periphery, maybe IR beacons or PING looking for specific posts or something like that. GPS may end up being of little use.

    What if I use the Javalin Stamp? I have one on order and I like the programming structure better. As far as the accuracy of the chip -- Is the accuracy off each time the unit gets a position or Is the unit off for all points during a certain time period. Example would be that today the temperature is hotter and it's cloudy so the reading during an hour period today is off based on the same locations the day before where the atmospheric conditions were different? Am I being clear on what I'm asking?
  • chunnchunn Posts: 6
    edited 2007-08-17 16:58
    Jim Richey said...
    Hey,Chunn

    Glad to see you're thinking.That's a challenging project to accomplish.
    I thought you might like to see a couple shots of my remote control lawnmower.
    It's been in full use for 3 years,I don't know how I ever managed to cut the
    lawn with out it as my yard is very steep and dangerous to even walk on.
    This machine,even with fully functional controls,has been upside down several times,
    on it's side,it's even tumbled end over end [noparse][[/noparse]that's the reason for the wheelie and roll bars].
    It would seem dangerous beyond imagination to allow something like this to roam around
    with nothing controlling it except for a puny silicon brain.Lawnmowers especially are
    capable of doing damage and destruction at a moments notice.

    Be Careful!! Jim Richey

    Jim,

    That looks awesome.

    Michael
  • chunnchunn Posts: 6
    edited 2007-08-17 17:02
    bobledoux said...
    I'm reluctant to reply because I'm not certain you understand the ramifications of your project.

    You have a multi-horsepower lawn mower, a very dangerous machine, operating without direct human control. What happens if it encounters an unexpected object, like a small child? Does it just run over the object? Your brief question doesn't give me confidence you are considering this fundamental issue.

    There is a good reason robot mowers use a random movement within a limited area of operation. That's because there is no simple way to provide path control required by a mower. Straight GPS is too slow and inaccurate to provide constant path control. It may be simple to navigate a robot from point A to point B, but a mower also must monitor the path between the two points to achieve the mowing pattern. Achieving that goal is not a trivial pursuit.

    Here is one approach: http://www.landscapemanagement.net/landscape/article/articleDetail.jsp?id=169360

    bobledoux,

    You are correct. I don't know a lot about the ramifications yet. I do plan on being very cautious because I have small dogs and small children. The unit will be on the riding mower and I will be riding along with it for a long time. Even after that I plan on having a remote control disable so that I can kill it remotely.

    I do plan on using some type of random movement. I only want the GPS to keep me in the boundaries even it it means 20 foot inside the polygon defined. I have 9 acres to mow to 20 foot inside is not much to spare if needed.

    Even if this does not work it's fun to do. What if it does work???

    Thanks,

    Michael
  • Jim RicheyJim Richey Posts: 82
    edited 2007-08-18 11:22
    bobledoux,

    The reason for the 2- 12 volt batteries connected in series is that the drive motor is originally from
    a handicapped scooter which requires at least 18 volts in order for the electric
    brake to operate.The motor,by itself,would run fine on a single 12 volt battery but the brake is
    needed quite often in order to to prevent a crash.

    Jim Richey

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Thanks, Parallax!
Sign In or Register to comment.