Is there a list of companies available that offer node to node distance calculation based on Decawave?
Most I find are local location based systems requiring anchors.
We have a point to point range only system, it’s effectively a single anchor variant of our positioning system.
(Hopefully obvious disclaimer, I work for racelogic)
I’ve never actually looked at the web page for it before, looks like they over did the copy and paste from the position system. This doesn’t give the speed and position as an output, it gives the range.
I should warn you that this is intended as a commercial/industrial system and is priced accordingly. I don’t want to discourage you if it does what you want but I don’t want you to get sticker shock either.
I’m sure that Humantics used to offer something similar but it seems to have vanished from their web site.
Andy,
Do you do multi-point or just single point to point? If you can, how many nodes can range to each other when in wireless range of each other?
The current firmware / software is up to 5 nodes each measuring range to a single point.
Each one can either range to one of the other 4 units or to some other non-measuring node.
So you can have several nodes measuring to the same point but a given node can’t measure to multiple points.
The underlying protocols are capable of doing multi point measurements but our original use case didn’t require it and it added complexity to the setup and configuration process so we left it out to keep things simpler for the user.
What would your use case be?
Andy,
What I need is ad-hoc ranging between mobile items. Each node will not be able to know before hand the ID of other nodes coming into the area.
There may be up to 32 mobile items that can come in and out of the area.
The application does not support use of fixed anchor for coordination.
As it stands our current firmware wouldn’t be suitable for that application.
The primary goal of the system as for high speed, high accuracy, low latency tracking of a single item with high dynamics. So we compromised the flexibility in order to increase the update rate.
But assuming you don’t need a high update rate I don’t think it would be much of a change to implement what you are looking for.
As part of the start up process the tag will send out a roll call message, all beacons in range reply, the tag then measures its range to each of them and then queries the closest for the system configuration it should be using.
If we were to loop that basic process and output the ranges measured it sounds like it would meet your requirements. We’d need to implement a method to avoid conflicts as things move in and out of range of each other but that shouldn’t be too tricky, we already do it for multiple ranging systems.
However none of this process is very efficient from a time perspective, it was intended as a one time startup process so the update rate would be fairly low.
Very roughly what would your required update rate be and what would the approximate budget be?
e.g. if you needed 100 Hz and $1,000 total cost then it’s not going to work. If you needed 1 Hz and the budget was $100,000 then we could probably come up with something custom for you.
If this is worth pursuing any more then please private message me.
All beacons need to range with each other and update needs to be 300ms or less.
There is no fixed infrastructure (ie. No anchors, only devices on mobile items that can come in vicinity of each other)
If I could use anchors or could determine which units are in vicinity then I could solve this relatively easily.
I have some other solution options that do not use Decawave that I may need to look into if this gets too involved for this application.
32 nodes, each measuring their range to up to 31 other nodes in 300 ms works out as 3,307 range measurements per second.
I’ve not seen any TWR style system that can reach that rate.
There are likely only to be 2 to 5 units in area at any one time but total number of say 32.
This means the ID of the few items in the area can be any of the 32 items that can enter the area
But if you don’t know which IDs they are you need some method to avoid collisions between multiple replies.
I think it’s doable assuming you don’t mind the update rate being variable depending on the number of nodes and the possibility of the rate stalling for brief periods if interference causes packet loss at the wrong time so the nodes can’t work out whos turn it is.
I could put a single beacon with long range capability to act as a coordinator and have each node with a receiver so that it looks for other items in a particular slot but if anything happens to the beacon the system goes down.
Another idea was that each item sets itself as a coordinator if no other items are found then when more items are detected in the zone, they form a dynamic network initially then all range with each other as a group until items leave and the network shrinks.
Each item only needs to detect up to around 15m so if the nodes had range up to say 25m then this self organising dynamic network could be established before they get within the range I am interested.
I think I could build something that would do what you need based on our hardware but with special firmware. But it would be a fairly substantial change to the structure of the firmware and radio protocol, beyond the scope of a quick hack.
So I think we could possibly do something for you but it would come down to budget and commitment, I don’t get the final say on development priorities but I suspect it’s too much work for us to put something together speculatively.