But, I don't know how to create all code in visuino for mover robot, because only arduino code are aviable from internet now. And think is better have Arduino code.
I watched video where some peoples connect two arduinos together, one only for GPS, and other main code for all rest.
( Arduino GPS Robot - YouTube )
So, can I create this:
-first arduino board - Arduino Main mower code
-second arduino board- Visuino code GPS ?
Visuino code show "outside" when mower is out the border, and "inside" when is in area.
Is that to much work with code to just include that inside/outside sintax from visuino into arduino code?
GPS is subject to various factors that interfere with the signals from the satellites that causes the 3-5m limit on accuracy.
There are units that can talk to another GPS whose position is known to get adjustments based on where the fixed unit calculates its position. Those adjustments can get you accuracy down to 10cm, but the equipment is more expensive than plain GPS units.
As @wildbill has suggested do a search for differential GPS.
Whilst I work with GPS quite a lot, I have no experience with differential GPS, they are not simple to setup and not cheap.
As far as I know, none of the very many commercial robot lawnmowers, that have been available for 20 years or so, use GPSs to navigate, perhaps there is a reason for that ?
Your missing the point really, GPSs do loose comms with satellites and report no fix, but if the GPS starts reporting a position that is say 20m out how can you possibly know its 20m out ?
It sounds like that GPS uses multiple satellite constellations concurrently so it's plausible to suppose that it has better accuracy than its predecessors that didn't.
I suspect that that is best case however. The example I found for sale wasn't expensive - maybe get one and test it.
There is little difference in accuracy terms between the various standard GPSs. I have tried two models from the Ublox M8 range and found they are no more accurate than others when used in my garden.
Under ideal conditions, wide open spaces with a good view of the horizon, you can get better than 3m-5m accuracy most of the time, but not always.
And there is the problem, whilst you might get good accuracy a lot of the time (and could produce a video to prove it) you cannot guarantee 1m accuracy all of the time, which you really do need for a robot lawn mower.
One of the best GPS modules, in terms of power and fix time is the Ublox SAM M8Q but low cost they are not. That GPS does use multiple satellite constellations and the data sheet stated accuracy is;
All the video appears to be showing is that the short term position stability or repeatability of the GPS reported position is in the 1m region, which is about normal.
There is no indication as to how accurate the locations are to the real world positions at all so they could be way out.
To work out the real position accuracy you would need access to a surveying quality GPS or a GPS reference location.
I was trying to remember the name of a system that does give better accuracy and finally found it: Pozyx. Expensive though.
I don't know how differential GPS works under the covers but I'll guess that the fixed station has been told its precise lat long and broadcasts the difference between that and what it's getting as a satellite fix for the mobile unit to use. There may be more to it than that though because these systems command a pretty high price.
It might be interesting to get two identical GPS units and some radios and build your own version of this.
