I’m starting in the DWM1001 world, and been reading docs for almost a week now.
My question is: can the DWM1001 3D locate fast moving tags, let’s say at 120kmh? For example could it be used to track RC racing cars to be able to draw moving dots over a graphic track course showing the car’s postion in real time, sort of like in the F1 live info??
The fastest update rate for DWM1001 is 100ms. Assuming a speed of 120km/h, it means you could have a measurement every 3 meters. It this satisfy your requirement then it’s doable.
With this update rate, the tag consumption will be higher so you will need a sufficient battery.
Thanks Yves, battery is not a problem.
As for the 100ms rate, I’ve read those 100ms are related to the Superframe size. Would there be any way to modify the superframe size by reducing the number of TWRs (I need only 6 or 8) tags. Reducing the amount of TWRs in the superframe, could then the update rate be increased??
Each TWR takes 5ms, using for example only 6, we could reduce the Superframe time by 50ms, that’s half the size, thus double update rate??
Unfortunately it is not possible to modify the size of the superframe and 100ms is the fastest rate that can be achieve with the DWM1001 PANS software.
If you need something faster, you will have to develop a custom solution.
Finally is it a hardware or software limitation?
Checking the code samples for the DWM1001, I’ve seen this
/* Maximum count of nodes in location data */
#define DWM_LOC_CNT_MAX 15
in the dwm_api.h include file, I guess that defines the 15 TWR in the superframe…
just kindda thinking aloud…
Reducing the number of TWRs I belileve would require a firmware modification, right?
It would require extensive software modification in order to reduce the number of TWR slots. The sole modification on this define will not change the super frame structure.
With a custom made software stack and UWB frame, it should be possible to increase the update rate for the same hardware.
Cool, so there’s still hope
Any idea of the improvement range we’re talking?
Thanks a lot Yves
Well it’s hard to determine, but out of my mind i’d say you could double the performance. The current super frame is 100ms and contains 15 TWR slots of 5ms each.
So reducing the number of TWR slots to 5, you could save 50ms, and double the update rate.
Keep in mind that it is not possible to modify the PANS software provided by Decawave. You would have to develop a full custom software stack which is quite an amount of work.
If I was you I would have a look at why you want UWB. if you want to track moving parts/vehicles GPS or accurate GPS (for example ublox NEO-M8T) might be a better fix.
If you really need UWB (indoor) you might want to add a software layer that also accounts for accelerator data and current speed vector to guestimate the result in the meantime.
Hi…You could use a bigger sensor or have a number of sensors all together, increasing in size. You would have to have some logic to tell you what you want to do when a particular sensor hits a wall but a smaller sensor doesn’t - and, of course, don’t count larger collisions when smaller sensors collide.The more obvious option is to increase the rate at which physics calculations are performed, though there is a processor penalty for this. I found it useful when I was moving a platform around which needed stacks of objects on top of it to move without going crazy.
quick-turn pcb assembly