I am reading tag data at 10Hz continuously for several hours. 10Hz sampling rate is set for stationary scenario as well. I am using UART channel to get location data from tag using dwm_pos_get() API.
Has anyone observed the tag reporting same data for multiple reads (around 30 reads spanning ~3 seconds) even though the tag location is changed?
Is there a possibility that the UWB network becomes non responsive for a while that all the tags just provide the same position based on the last successful TWR until they get an updated location info based on a fresh TWR?
FYI, I am using less than 15 tags in the network.
Another observation is that the change in tag data is very slow at times compared to the speed at which it is moved.
Any insight in to such observations would be of great help.
what you see is an expected behavior, tag uses two update rates when stationary detection is enabled. When you move with the tag, the update rate changes from stationary (default=100) to nominal (default=1). When you call dwm_pos_get() between the updates, you get that lates position/distances, so there is no need to read position more than once per update period. It is not that the UWB network is not responsive, it is that you have high update rate and you are reading faster than the position/distances are being measured.
You can disable the stationary detection to use only the nominal update rate or you can change the update rates using aurs shell cmd or dwm_upd_rate_set().
Though I have enabled “STATIONARY DETECTION”, I have set the update rate to 100msec for both normal and stationary modes. Also, I am calling dwm_get_pos() every 100msec. In case of any mismatch between my calling rate and the tag update rate, I understand that the location data can repeat twice at the max once in a while. The observation here is that at times I see many consecutive repeats (easily 15 to 20 reads with same value). This has been fairly reproducible. As I use multiple tags, on one particular occasion, I have seen all tags repeating same value for close to 3 seconds. That’s what led to my suspicion if the UWB network went down briefly.
Anyway, as suggested by you, let me try out after disabling “STATIONARY DETECTION” altogether and then update my observations.
After disabling STATIONARY DETECTION, setting normal update rate to 10Hz and reading tag data every 100 msec, I continue to observe same position data repeating with a moving tag occasionally. At once, I have seen 60 consecutive reads of same value spanning ~6 seconds.
This is a major blocker for our product due to such reliability issue.
May I request someone from DWM1001 design team respond to such possibility?
I would recommend to use interrupt event to receive the updated location every time it happens. This is a reliable way to not oversample/undersample the location and keep the module in sleep mode when no new data arrives.
Just enable the interrupt and let the module wakes up the host MCU/UserApp when it has new data. Then put the module to sleep after data is processed. See the documentation for more details.
another possibility is that the tag and the anchor don’t see each other properly, in that case, the measurement is not possible and you will experience what you describe, try to put the tag and the anchor close to each other and try again to rule it out, otherwise the measurements should work as expected and you should get new values every update period
Yes. When I cover the tags with my hands and read the data, I could observe that the position data repeats along with qf=0 probably indicating that the TWR has failed or unreliable.
In the actual set-up, the tag is fairly exposed to anchors during movement. Also, data repeating for about 60 reads spanning 6 seconds looks even more strange. So, I am not able to relate to the possibility of tag unable to see the anchors continuously for that long.
Let me do some more experiments in similar lines to what you suspect and check if I can get any further clue. Thanks again for sharing your thoughts.