Delay for dwm_loc_get command

I have a four anchor one tag setup and I call dwm_loc_get in C. It gives me a position and timestamp. What is the delay on this data from the real-world? Does that give me the latest timestamp that was actually calculated 1 or 2 seconds ago? Or does it try and calculate the position right when I ask for it?

I’m asking because we see our UWB data is delayed around 1-2 seconds when comparing it to ground truth. So I’m trying to figure out if the problem is in the UWB data collection.

Thank you!

The issue is that the decawave system is, like most UWB tracking systems, designed for slow moving items with slow update rates. Most people don’t care too much if their position information is a second or two old.

If you need accurate time information for something that is moving at any sort of speed (and if it isn’t moving much the time accuracy doesn’t matter) then to get good results the system needs to be designed with that in mind from the start. The more accurate the position you want to get from it the lower the speed where this sort of thing starts to matter.

Do you see what looks like slightly different output delays and time errors depending on the direction of travel?
I believe the system you’re using is using a group range method, each anchor replies in turn to the message from the tag. But they aren’t replying simultaneously, that’s simply not possible for UWB radios. At least not for these UWB radios. Which means the range to each anchor is actually measured at a very slightly different time. If you are moving slowly enough then this doesn’t matter, it gets lost in the noise. Once you start moving quickly this can start to skew your position calculation unless you take it into account. If the anchors are always measured in the same order then you will end up seeing a position bias that depends on the direction of travel. If you try to line things up based on position against a reference then this will look like a varying time bias.

None of which actually answers your question. But hopefully helps explain why the results you are getting aren’t particularly good, you’re asking it to do something it wasn’t designed for.

Cheeky bit of self promotion:
If you want I can sell you a UWB system that outputs positions timestamped to GPS/UTC time with a sub ms accuracy. :slight_smile:

Thanks Andy. I don’t care about the delay I just wish they gave me a timestamp that I could line up with my other data.

I open to learning about your UWB system, but I’m in a GPS denied environment.

Hi @esteimle
the PANS system is operating in TDMA scheme so each TN have reserved its time slot where it can do ranging with surrounding anchors. The maximum TN update rate is 10Hz (100msec) and the output value is an average of last three positions (this could cause delay between reported position and real position if the TN is moving fast). But you need to double check update rate settings and responsive mode settings - if you are using default setting then when the TN is not moving it will fall into slower update rate (1sec). Once it will start moving again the node switch to faster update rate - this could also cause some delays.