I’m very new at this whole GPS thing and I’ve decided to challenge myself to a pretty complex project, so any help is more that welcome.
I’m looking to build a robot that can help me find specific points in a field, down to about 5 cm accuracy is needed. The field is flat, cut grass (so not extreme) and is generally a good distance from buildings (over 50 feet at least).
What boards would help me get started with this?
I’ve purchased this board but am having a real struggle to get it to actually acquire centimeter level accuracy, the closest it gets is DGPS (about 1 meter). I’m ready to throw the towel at it blaming it’s tricky documentation and possibly because it was cheap.
I’m looking to keep it simple-ish, I don’t need it to travel fast but I do need it to be accurate. I am currently looking at this:
I’m not sure if this is the best fit for the job.
Overall:
4 cm accuracy (doesn’t have to be finer than that) RTK?
wifi will be provided by my phone hotspot or bluetooth if that’s an option
I’m currently targeting Wisconsin, so there’s a solid base-station setup here that I’m already signed up for and using
I’d like to learn but also not get frustrated, so anything to help move this along and get the “yep ok I can get the GPS thing down, now to make that move a robot”.
I would like to see it’s information via my phone, I’m willing to create an app if needed and maybe just bluetooth information from it would be good enough. I plan on having a custom app eventually that’ll help control the robot and see information about it.
Would I need a different antenna? The one that came with the waveshare is small, black and square… I put metal under it to help the signal but didn’t seem to help much.
Also if anyone has any advice on how to get this to the next step of moving a robot. I’m thinking a robotic lawn-mower sized thing, two wheels for driving and pivoting, one caster style wheel in the front.
Any help would be awesome, I feel I’ve hit a wall and am willing to spend a bit to get this project off the ground before it starts really getting nice outside.
^As above…if you’re using a Pi you’ll probably be happy
In the meantime:
Here’s a video that shows an example of the postcard’s tracking via SWmaps - you’d add your Wisconsin correction info and pair your phone and then they’ll pass that to/from one another (if the postcard turns out to be viable for your use-case, if not we have other options that work very similarly!)
I haven’t even thought that far ahead yet, I was thinking once I get data from the GPS portion of the bot that I’d start working on using that to help drive the motors/wheels/etc. lots of “ok future maths of I am here, I want to get over there, turn… go… etc”
Since I want to make this easier on myself, I’m very interested in “plays nice” since that almost implies there’s ways that it won’t I’m up for suggestions for sure.
For the overall goal, is I basically need this robot to find corners of a square (large square, think 300ft type of square). I was planning on taking it to a proposed corner of the square, then push a button either on the phone or bot to say “this is the corner”. Then walk some distance away from that and do the same “this is a side of the box”. Then set it down and let it do it’s thing. I’m actually ok at that point to mount the phone on the robot so that it basically has it’s wifi/cellular/bluetooth just kind of there for it as it works.
Thank you, yes that video is actually what drove me to think the postcard was a good option. It looked so easy to get CM level accuracy, and she’s within sight of a huge metal building… so it’s pretty good at finding that accuracy without much hassle.
That video inspired me to say “ok what if it spoke bluetooth to my phone for helping setup” and “use that with say a raspberry pi or other robot driving board to help tell it where to go, turn, move, etc”
