We have been experiencing chronic problems using LPS when working with a set-up with the following HW/SW configurations;
- Crazyflie 2.1
- Flow deck v2
- LPS installed with 8 anchors on TWR mode according to https://wiki.bitcraze.io/doc:lps:twr-reference-setup
- An example beacon set-up (from among many others we changed) can be seen in attached lps1.jpg
- We are using the autonomousSequence.py script giving different set-points to see some self flying behaviour (later some tracking + path planning will be integrated)
When CFclient is connected with Crazyflie on the ground facing the appropriate x direction; None of the anchors located at ground level is identified as seen in image attached lps2.jpg.
This changes however if we place the Crazyflie at its same starting position but on an elevated level i.e a chair or box. In this case most of the anchors are seen.
Some Approaches we considered
- We have alternated the positioning in various ways; placing all the beacons on the ground on a higher level (i.e. each on a small box), we placed beacons slightly closer to each other (not closer than 2m as adviced in documentation). But the same issue recurred. Is there anything you may point out on these?
- There were explanations on max range at https://wiki.bitcraze.io/misc:investiga ... -max-range. I assume this suggests changing the chip settings of the beacons in some way?
- The area we are using for the Crazyflie is a robot soccer field of 12m x 9m dimensions which is inside a departmental area. But beacons are placed closer than the field size so connection is stable. Yet could indoor room size we use have an effect on beacon communication?
Yet again, we would like to get your suggestions on how to obtain full coverage from our beacons.