Hi
In our project, we have GPS isuess to filter the GPS location, we need ML to choose the correct location for MCU to report to the cloud.
Can edge impluse do it ?
Hi @jian, interesting question. Could you elaborate a bit on the type of data that you have, and what the typical error is?
1 Like
Our device is uing for animals and tracking their location.
We have problem to get the correct location from animals.
GPS drift is big problem. We also using IMU sensors to give a speed for animals.
Do you have any idea how does the tinyML could determine their location ?
Can you send us a picture of the GPS tracker and the name and manufacturer. I have found the position of the antenna on the animal makes a big difference. 8 second average TTFF when there is a good view of the sky.
Hi
we are using https://wiki.dragino.com/index.php?title=Lora/GPS_Shield
which the modules is https://www.quectel.com/product/l80.htm
the antenna is on top of the animal’s body.
@Kimwo, good suggestions!
@jian, could you share some of the faulty data, and what the correction would be?
OK I have played extensively with the L80. Although the chip shown is quite close to other components (contrary to the design guide) it should work well. However it does take time to collect the ephemeris data and become accurate. So 1) are you leaving the GPS chip with power to v_bckp pin so that it does not have to start from scratch each time? 2) does it take the first fix it gets - the accuracy increases with the second and subsequent fix. It looks like you can connect directly to the L80 and read the data. In which connecting something like PowerGPS will show you what is going on. I would remove the lora board while testing as shown in the manual. Have fun.
2 Likes