How to locate a person in a field with UWB?

I’ve been trying to figure out what fits my scenario from the salad bowl of acronyms (AoA, AoD, PDoA, TDoA, TWR etc. you get the idea), and so far have failed; AoA with TWR seem most likely.

This is the hypothetical scenario: A group of fire fighters parachute out of a plane into a smoke filed field. They are close enough to each other that whatever tech will have enough range in this hypnotically scenario, there are no obstacles in the field (to keep this simple). They can not see each other due to the smoke.

Using UWB how do I know the bearing and range (without GPS, I’ll get to that) to the nearest colleague? That colleague could be injured and not moving, or running because of the fire. Could be running toward me or away from me or any other angle relevant to me.

I can currently do this with GPS on each unit and transmitting the Lat/Lon. A compass gives me a ground reference to set the zero point for the bearing. An arrow points in the direction, via 0->359 degrees in circle, I need to go. I know the range by the standard distance calculations.

Now for whatever paranoid reason you choose, GPS is no more.

How do I use UWB to do this with things that are available today or in some reasonable amount of time (Say before Q4 given the state of the supply chains)?

Most everything I’m finding is a structured environment like a warehouse. My application is totally unstructured environment. I don’t want to be dropping breadcrumb way-points and if I did they would be randomly placed in X/Y/Z space due to time constraints (seconds).

Can’t use GPS? Use Galileo. Or Glonass. Or Beidou :slight_smile:

AoA (probably calculated based on PDoA) and TWR will in theory give you a bearing and distance between two objects. As such it would in theory do what you want. You may however have issues with their bodies blocking the signals.

Assuming a reasonable number of people (>3) a pure TWR system may also be an option, you could create a mesh of ranges between each person and from that calculate their relative locations. Body blocking is still an issue but the simpler antenna requirements give you more options to work around this.
This would however have some ambiguity, you could get a solution that was mirror image of truth. To resolve this (assuming it matters) you would either need some angle measurements or basic inertial integration and track items over time.

Thank you.

Does any such hardware chips or modules exist today that do PDoA with TWR?
I know some of the chips claim to support such things. Have not found actual data sheets for them. The Rep tell me that is intentional.

I know from the form here that PDoA is the focus Deca/Q is taking. Have not seen anything like a release date for any actual hardware.

The DW3000 chip has a variant that supports PDoA. However there are currently no off the shelf modules that I know of that include it, they are all for the single antenna version. So you would be looking at building your own board with the chip and suitable antenna array.

The Boss insists on modules to avoid having to do our own FCC certification.

If I use two certified antennas am I good?
I don’t think so, not sure.

In that case you are limited to the DWM1001 running the PANS system firmware which is TWR based and requires anchors in fixed known locations.

The radio protocol used is part of the certification. For things like WiFi they can bake that protocol into the module and so the module is always certified. Most of the time with UWB you are running your own protocol and so need to pass certification.

If you must use pre-certified modules then maybe look at some of the bluetooth based ranging options out there. They aren’t as accurate but get around this issue.

Or put an iphone on each person and write software to range between them using the protocol Apple have already certified. :slight_smile: