monkeyfork
Posts: 62
Joined: Tue Oct 29, 2013 10:14 am
Location: orlando

Re: catPooPee

Tue Apr 02, 2019 4:12 pm

PhatFil, thanks for help. The links you provided gave provided a whole new dimension for my project (cat tracking). This code base is especially interesting https://github.com/switchdoclabs/iBeacon-Scanner. Most of my code for work is in C but much much easier these days to find examples for hobby stuff in python I guess. I'm just at the dangerous skill level with python and it always annoys me because of the;
1) 2.7 vs 3 stuff
and the
2) how can there be Sooooo many ways to add modules stuff
However, the code base looks managible. I tried getting that going yesterday but ran into numerous problems. I'll take it to my local linux user's group and see if i can recruit a cat-person.

Today, I'm maintaining sanity by NOT doing software. Rather I'm working on gridPee, the gridded urine containment vessel that is a short drop and a sudden stop under the fake reuseable "dirt" (safflower seed) that the cats scratch in then do their business. I'm trying to make something that isolates the drippings within a few grids so as to facilitate a dipstick check or a pipette sample. I'm using off the shelf stuff like a grid that's a light louvre sealed to a plane of plexiglas with cheap bondo epoxy. Here's a pix
Image

Cheap epoxy outgasses toxicity for years so this stage is just proof of concept. Maybe I can move to a 3d-printed grid that can optimize grid geometry vs standard-cat-pee-volume.

monkeyfork
Posts: 62
Joined: Tue Oct 29, 2013 10:14 am
Location: orlando

Re: catPooPee

Wed Apr 03, 2019 9:59 am

@Andyroo @PhatFil concerning BLE-beacon ranging for triliteration, it's turning into a very shiny object. I've found this paper https://www.semanticscholar.org/paper/ ... ce36e0336 that gets to reasonable accuracy thru channel diversity and more importantly Kalman filtering. Kalman is magic, I've seen it clean up noisy nav and optical sensors in enormously expensive aircraft systems. I like the approach that the authors of the paper employ with the channel diversity to attack multipath and fast fading (to tighten the RSSI accuracy) but that would require extensive coding of the Nordic nrf51822 chips used in the cheapie BLE-beacon tags that I want to use, my pensioner budget doesn't allow purchase of dev-tools to flash those chips. However, (BIG IDEA ABOUT TO BE REVEALED) throwing in another rpi-zero-W to gain another BT receive antenna (and RSSI digitizer) in close proximity to the main catPooPee processor could possibly have a similar effect on accuracy. I'll be able to make a first pass look at this idea in about a week, stay tuned.

monkeyfork
Posts: 62
Joined: Tue Oct 29, 2013 10:14 am
Location: orlando

Re: catPooPee

Thu Apr 04, 2019 7:36 am

Concerning BLE-beacon tracking to determine which cat is in the litter-box/tray. I've just found there is an open code base for the Nordic chips used in BLE-beacons https://github.com/meriac/openbeacon-ng . This greatly increases the chances that my project might feature-creep into the world of generalized indoor-cat tracking if openbeacon will port nicely to rpi3 or rpi-zero-W. If anyone has been down this path before me, it certainly would be nice to know. As mentioned before, I am ancient, I do not need to be re-inventing wheels or tilting at windmills.
regards.

monkeyfork
Posts: 62
Joined: Tue Oct 29, 2013 10:14 am
Location: orlando

Re: catPooPee

Fri Apr 05, 2019 3:13 pm

The under litter gridded urine containment system (really need a name that works out to funny acronym) has been epoxy sealed now and first test performed. approx 0.5oz of green dyed water was poured through the non-absorbing litter (safflower seeds) and resulted in a pattern of grids being filled and overflowed to some adjacent grids.
Image.

Interesting observation of watching the motion video of my furry lab assistants. Piffy(female) pees like a laser. PeachO(male) pees like blunderbus scattergun... no doubt trying to write his name and being hindered by his poor cursive skills and unfortunate spelling confusions.

Great progress yesterday receiving BLE-beacon packets from a simulated BLE eddystone packet on the intended rpi-zero-W target. Now feeling confident that BLE-beacon might be a very good tech for openCatTracking. I'm still kind of mystified why it is necessary to use a quite convoluted set of blueZ utilities to get the necessary information for position location... almost like the obfuscation is by design.

Andyroo
Posts: 3373
Joined: Sat Jun 16, 2018 12:49 am
Location: Lincs U.K.

Re: catPooPee

Tue Apr 09, 2019 12:37 am

Just watching https://youtu.be/y89yJ1Fq-hQ

It’s RF-ID based with three different readers using either the under skin or ear tag based chips.


Spoiler alert:

Main issue is range and orientation- 90deg to reader gives zero read! Even with perfect orientation the first module only reaches 5cm with the ear tag while the long range reader manages 11cm with the under skin tag and 20cm for the ear tag.
The best result is by using a ring that the cat goes through for food...
Need Pi spray - these things are breeding in my house...

monkeyfork
Posts: 62
Joined: Tue Oct 29, 2013 10:14 am
Location: orlando

Re: catPooPee

Tue Apr 09, 2019 10:42 pm

thanks Andyroo, the high energy reader looks like it's worth a try. (btw, the swiss guy is hilarious). Seems to me I could fab a loop of same inductance as their large loop that could run the perimeter of the litter box just above cat-shoulder height and have fair chance of detection. I'll try to contact the swiss guy and see what he thinks.

I've been researching and experimenting BLE-beacons. I'm trying different configurations to try to understand why the RSSI numbers I see are so gawd-awful. One experiment was to shutdown WIFI on a rpi-zero-W and collect data only thru the console serial, so that there wouldn't be any WIFI transmission when the RSSI was being measured. With a small data set that appeared to reduce the spread of numbers down to around 10 (is that really dbm or some relative scale that relates to nothing but the whim of a Nordic marketeer?) ... who knows, documentation in the bluetooth world totally sucks). Anyway, 10dbm variance for a non moving radiator and receiver is awful. I'm guessing the BLE stuff was never really intended to be used for location. Having said that I do have confidence in exotic filtering, we shall see.

Andyroo
Posts: 3373
Joined: Sat Jun 16, 2018 12:49 am
Location: Lincs U.K.

Re: catPooPee

Tue Apr 09, 2019 11:11 pm

Two main issues that may be a problem.

The first is the inductance of the loop. Having to get a large number of turns to act as a power source but keeping the inductance right so you do not create standing wave feedback or blowing the input stage of the receiver (if I remember my radio days).

The second may be the real problem. I think the animal tags are under the skin along the back? If so, will this be parallel to the loop? If so range and read ability will drop...

One other thought is to link a ESP32-CAM and go for vision recognition?

I’ll keep my eyes open for you though.
Need Pi spray - these things are breeding in my house...

PhatFil
Posts: 1242
Joined: Thu Apr 13, 2017 3:55 pm
Location: Oxford UK

Re: catPooPee

Tue Apr 09, 2019 11:57 pm

So sorry it took me so long to think of this, but if you can get your hands on an original pet porte cat flap, you can actually reflash with an alternative firmware.. https://respekt-empire.de/CatFlap/trac
(Pet Porte cat flaps allow entry based on embedded rfid pet chip, all the hard work done for you in a commercial product.. )

Turn your litter tray into a litter box with a pet porte door to enter and you can id the cats via their rfid chips.. DOH!!

monkeyfork
Posts: 62
Joined: Tue Oct 29, 2013 10:14 am
Location: orlando

Re: catPooPee

Wed Apr 10, 2019 10:33 am

Thanks Andyroo and PhatFil for inputs. All will be evaluated.

My bad on diss'ing Nordic marketing earlier for the bad RSSI readings I'm seeing... I finally realized that in my current BLE-beacon experiments only broadcom chips are involved. I'm simulating a BLE-beacon with a RPi-3 and the receiver is a RPi-zero-W. So, on both ends of BLE-beacon experiments the BCM43438 (now CYW43438) is the hero or culprit. BLE-beacon for cat-proximity and possible cat-tracking testing will continue (woefully underfunded) as I drill down into the 43438 docs and flesh out the beacon network. Shortly BLE-beacons tags using Nordic NRF51822 will arrive and merge into the test plan.

meow

monkeyfork
Posts: 62
Joined: Tue Oct 29, 2013 10:14 am
Location: orlando

Re: catPooPee

Sun Apr 14, 2019 5:54 pm

Continuing with BLE-beacon research/testing to hopefully use RSSI to detect which cat is in the tray. I've read some academic and industry research that uses some NON-trivial filtering techniques that have promise but currently are not in my toolbox. Other methods for sharpening the range to target use frequency diversity (quick succession of beacon transmissions on different frequencies) to defeat noise. That method requires difficult software in the beacon which is beyond my patience/budget right now. Rather than pursuing those solutions I propose to experiment with multiple beacon receivers to better understand the nature of the noise that makes RSSI measurements so horrible. My primary receiver so far has been a rpi-zero-W because it is the primary choice for making my project easy to reproduce as open hardware/sw. I've just ordered an additional rpi-zero-W with the intention of taking simultaneous RSSI measurements (while I vary the geometry of the receiver placement). I can easily control the periodicity of the BLE-beacon transmission, so it will be a piece of cake to insure that I am recording simultaneous RSSI values at two receiving points. I'll be collecting data via serial and NOT-WIFI to factor out the chance interference of WIFI into the RSSI measurement. Watch here for data soon. If anyone has data on antenna pattern of WIFI/BT antenna on rpi-zero-W it would be of great interest.

Andyroo
Posts: 3373
Joined: Sat Jun 16, 2018 12:49 am
Location: Lincs U.K.

Re: catPooPee

Sun Apr 14, 2019 7:40 pm

Maybe some way of attenuating the signal so you only pick up the cat in the box?

Not sure you could physically do this on the zero as it’s built into the card but maybe software wise it’s possible?

As for the name of the grid, how about Pussy Urine Receptacle or pur for short? :oops: :lol:
Need Pi spray - these things are breeding in my house...

monkeyfork
Posts: 62
Joined: Tue Oct 29, 2013 10:14 am
Location: orlando

Re: catPooPee

Mon Apr 15, 2019 8:17 am

Andyroo, this paper https://www.hindawi.com/journals/misy/2016/8367638/ suggests that a BLE-beacon close to the ground (cat wearing BLE-beacon on collar perhaps) will have such a short range that it will naturally be only useful as a proximity detector. I had hoped that proximity detection could be a generalized case of longer distance cat tracking. However, the more I research the less hopeful I become. Obviously, indoor positioning systems are a tough nut to crack.

Andyroo
Posts: 3373
Joined: Sat Jun 16, 2018 12:49 am
Location: Lincs U.K.

Re: catPooPee

Mon Apr 15, 2019 7:54 pm

If it cheers you up trying to identify the line action of a pallet in a warehouse was also a real pain. So much so we looked at cameras and labels to track them :shock:

I had a quick look at live facial recognition using a Pi and OpenCV here but even with very different human faces the accuracy was not great so for similar cats I think that would be a dead end.

I did once look into getting WiFi across the local church yard to a friend (I had ISDN, he had dial up) and came across the Pringles directional antenna build. I wonder if you could use two highly directional antennas (antennae?) to cover the area and only accept the ‘cat x present’ if both have the signal?
Need Pi spray - these things are breeding in my house...

PhatFil
Posts: 1242
Joined: Thu Apr 13, 2017 3:55 pm
Location: Oxford UK

Re: catPooPee

Mon Apr 15, 2019 8:46 pm

Probably been considered or dismissed before but enclosing the litter tray within a fariday cage could provide a solution to identifying signals simply, a pi with an ethernet connection within the cage should only pick up device signals also within the cage?

Andyroo
Posts: 3373
Joined: Sat Jun 16, 2018 12:49 am
Location: Lincs U.K.

Re: catPooPee

Mon Apr 15, 2019 9:02 pm

PhatFil wrote:
Mon Apr 15, 2019 8:46 pm
Probably been considered or dismissed before but enclosing the litter tray within a fariday cage could provide a solution to identifying signals simply, a pi with an ethernet connection within the cage should only pick up device signals also within the cage?
Just reread the thread but could not see anything on this - you may have hit a solution if cats are happy to go in an enclosed box (I use ‘go’ here in both senses).
Need Pi spray - these things are breeding in my house...

monkeyfork
Posts: 62
Joined: Tue Oct 29, 2013 10:14 am
Location: orlando

Re: catPooPee

Tue Apr 16, 2019 1:20 pm

Yup Andyroo, hearing of your struggles re IPS with tracking pallets does cheer me up because adding to the civilization knowledge base on the concept seems a more worthy goal now. PhatFil, even though I like the elegance of my furry lab assistants leaving their loads inside a faraday shield I think I'll hold out a little hope that I can crack this BLE-beacon nut and provide the world with whole house cat tracking as a shiny spinoff of poop weighing.

These guys in China https://www.mdpi.com/1424-8220/18/9/2820 seem to demonstrate lately (2018) that with simple Kalman, trivial particle swarm optimization and the use of a secondary gateway path for dual receipt of BLE-beacon transmissions that accuracy of RSSI ranges can be reduced to less than a meter. The "gateway path" is where I was going with my proposed two rpi-zero-W experiment. Humm, China... you'd think things like this would be going on in, oh say... Cambridge???? (please wake up some of your brainiacs at Bletchly, it would be very embarassing if Florida-Man cracks this cat cipher first).
Last edited by monkeyfork on Tue Apr 16, 2019 6:54 pm, edited 1 time in total.

Andyroo
Posts: 3373
Joined: Sat Jun 16, 2018 12:49 am
Location: Lincs U.K.

Re: catPooPee

Tue Apr 16, 2019 3:05 pm

monkeyfork wrote:
Tue Apr 16, 2019 1:20 pm
... that with simple Kalman, trivial particle swarm optimization and the use of a secondary gateway path for dual receipt of BLE-beacon transmissions that accuracy of RSSI ranges can be reduced to less than a meter. ...
Are you quoting ‘The Man with the Swiss Accent’ in his native tongue here?

You lost me a bit after the word ‘simple’ to be honest :oops: :lol: :oops: I tried hunting down some of the text in a few links but could make no headway with them - my math stopped at parrots climbing ladders in their cages (sin/cos/tan etc) except for the odd delve into Knuth fundamental algorithms many years ago.

With my mind as it is currently you will find me on the side cheering you on if you ask for help with these...
Need Pi spray - these things are breeding in my house...

monkeyfork
Posts: 62
Joined: Tue Oct 29, 2013 10:14 am
Location: orlando

Re: catPooPee

Wed Apr 17, 2019 1:48 am

Andyroo, the words "simple" and "trivial" was with one of my ex-wives's sarcasm switch engaged. My point (disguised) is that the BLE-beacon location remains a tough nut but I think there is a glimmer of hope that it will be cracked by an enormously complicated software stack
running on consumer commodity hardware (as is the norm). I hope to show enough cat health benefits from behavior monitoring (and poo/pee weighing) that some folks who are at the intersection of the population circles marked "cat-lover" and "brainiac" will eventually help me to layer on those "trivial" pieces of code like Kalman-filter and "particle swarm optimization". (side note: I met Dr. Kalman years and years ago and marveled at his real world abilities to help others tame a rather unruly exotic aircraft system)

monkeyfork
Posts: 62
Joined: Tue Oct 29, 2013 10:14 am
Location: orlando

Re: catPooPee

Fri Apr 19, 2019 1:10 pm

short status 19apr2019

ble-beacon two receiver measurements: I have two rpi-zero-W's allocated for investigating awful RSSI variance. both will be receiving simulated ble-beacon packets from a rpi3 in adjacent room less than 10m distant. My thinking is to minimize minimize RF interference by not using WIFI on either rpi-zero during measurements and sanitizing local RF environment. My method of seeing the ble-beacon packets on the rpi-zero's is the following cli lines; (gleaned from multiple sources... sorry, can't reconstruct that crime with links right now)

Code: Select all

sudo su
btmon &
hcitool lescan --duplicates
 
This results in packets being scrolled on the tty where the above was executed that look something like this;

> HCI Event: LE Meta Event (0x3e) plen 32 [hci0] 132.683171
LE Advertising Report (0x02)
Num reports: 1
Event type: Non connectable undirected - ADV_NONCONN_IND (0x03)
Address type: Public (0x00)
Address: B8:27:EB:B1:47:F8 (Raspberry Pi Foundation)
Data length: 20
Flags: 0x06
LE General Discoverable Mode
BR/EDR Not Supported
16-bit Service UUIDs (complete): 1 entry
Google (0xfeaa)
Service Data (UUID 0xfeaa): 100002706966667907
RSSI: -72 dBm (0xb8)


(if anyone could help me get the above reports via python or especially C , or even a link to forum thread
where such a goal is being pursued, I would be very grateful)


Notice 1st line of report, the [hci0] 132.683171 appears to be a time stamp, ie good news. The bad news is that the time is
relative to when "btmon" was started rather than related to system time. So comparing packet reports (from 2 receivers) by time stamp is only
as good as how quickly I can start btmon on two machines (so far that's about a half second of confused keyboard clicks
and mouse button hits). The reason for trying to sync this process is to aid comparison of two received packet RSSI's on the exact
same packet. So far, with some confusion, it appears that two receivers measuring the exact same packet in two different rpi-zero-W's
continues to give two wildly different RSSI's. The confusion in sync'ing this process will be eliminated and a data set of dual RSSI's will be collected. First step will be to shut down WIFI and have both rpi-zero-W's comm their results via serial to a intermediate collection host that also will not have WIFI active.


gridPee "gridPee" (aka P.U.R.) is gridded urine collection in lower portion of my prototyped litter box (tray) that is intended to make urine tests via glucose-dipstick or pipette collection easier for a non professional. I have started the process of transitioning my furry lab assistants to the gridded paradigm. So now, when they scratch to find the puurrrfect spot to do their business they will find a plastic grid (approx 1cm square) rather than the smooth floor. Filling the grid will be either a traditional clay litter or the safflower-seed litter-NT (new technology, ha).
They have already made the transition to the safflower seed (re-usable, non-absorbing) with lots of sniffing but no complaints. btw, the clay to safflower transition was made gradually as suggested by other pee/poo litter scientists.

remember to discuss catPooPee at the dinner table during Easter feasts :D .

monkeyfork
Posts: 62
Joined: Tue Oct 29, 2013 10:14 am
Location: orlando

Re: catPooPee

Sun Apr 21, 2019 1:54 pm

i made 1st data capture of 2 ble-beacon receivers (both rpi-zero-W) that were side by side oriented the same. receipt of ble-beacon packet was performed using new-to-me code set (python aioblescan). Both rpi-zero's we being operating from minicom sessions on a netbook that was placed far enough away from zero's to hopefully keep switcher noise away from zeros and comm'd to lan via wired ethernet (ie was not radiating on WIFI). Neither zero had any _INTENDED_ wifi activity (ie no ssh sessions but could have things like a periodic NTP packet flying). The screenshot embedded below was taken on another pc on the LAN but distant from the data collection point (so my RF perturbing bubbling salt sack would not absorb or multipath BLE packets). (screen shot pc is on a chromebook running linux via "crouton")
Image
First observations;
yes - natural sync of receiving individual packets on two side by side receivers is extremely interesting.
yes - the RSSI variation _seems_ to track on both usually within a few dbm (but i reserve the right to call BS on myself)
no- they do not track exactly and there are incidences of huge variation in RSSI from one zero to the other on the same packet.
no- packets are NOT always received by BOTH even though in same near field and no where near RF noise floor WHY WHY WHY?!
yes- this test scenario will be refined because their is more POO here than is supplied by my furry lab assistants and it MUST be understood.
Last edited by monkeyfork on Sun Apr 21, 2019 10:24 pm, edited 1 time in total.

Andyroo
Posts: 3373
Joined: Sat Jun 16, 2018 12:49 am
Location: Lincs U.K.

Re: catPooPee

Sun Apr 21, 2019 2:04 pm

packets are NOT always received by BOTH even though in same near field and no where near RF noise floor WHY WHY WHY
That’s odd - noise is the obvious thought being chased by duff software.

This may help shed a bit of light on the missing packets https://news.ycombinator.com/item?id=14752719
Need Pi spray - these things are breeding in my house...

monkeyfork
Posts: 62
Joined: Tue Oct 29, 2013 10:14 am
Location: orlando

Re: catPooPee

Sun Apr 21, 2019 4:39 pm

Thanks Andyroo for that link, finally some правда about bluetooth (and if you don't get the irony of writing 'truth' in Cyrillic you don't know any Russians). I like the guy's comment about layer1 spec
"like a mating dance between scorpions in the middle of a freeway"
. I wouldn't necessarily be quick to blame BT-oddities on the software stack though (at least for the more simple BLE and the unconnected beacon stuff). I'm leaning more toward some of the quick-2-market silicon on the cheap. I want to know a lot more about the broadcom (now cyprus) CYW43438 combo WIFI/BT chip. It's obvious that they multiplex the TX function between WIFI and BT. Usually, especially w/ consumer grade silicon, it would be hard to RX while TX'ing. So if a BT packet comes along while WIFI is TX'ing no way the BT packet is going to get received. Soooo... if you have any insight into how I can assure that WIFI portion of the CYW43438 is totally inhibited (better yet powered down) while leaving the BT functionality active the details would be hugely useful.

GridPee increment My furry lab assistants (PeachO and Piffy) are now exclusively prancing/scratching on safflower seeds as they prepare to poo/pee. They seem to walk very lightly and gingerly, quite the odd show. The ministry of silly walks is probably because of the sugar-sand effect of the safflower seed, ie they sink quite a bit more before compaction and seed-lock(tm) give them firm footing. (cats are no doubt replaying old movies in their mind where good guys ALWAYs find themselves mired in quicksand). Below the generous mounds of safflower seed (that can be re-used) is the plastic grid (re-purposed lighting louvre) filled with traditional clay litter (that can NOT be reused). This odd stack in the litter box (tray) is just an interim step to indoc the cats and watch behaviors (don't want to run risk of pissing off my pissers by changing paradigm to quickly).

monkeyfork
Posts: 62
Joined: Tue Oct 29, 2013 10:14 am
Location: orlando

Re: catPooPee

Tue Apr 23, 2019 2:11 pm

there will be a short pause in cat-tracking feature creep

short-explanation with background. I spent centuries (it seems) watching brainiacs at Texas Instruments Defense Group (now Raytheon) do miraculous things like;
monopulse resolution improvement
surface acoustic wave FM chirp discriminators
inverted synthetic aperture radar
gold code detection of first GPS signals
precisely pointing antennae with no moving parts
you know, trivial stuff that takes crappy signals and makes sense of them by the application of scary wicked intellect.

fast forward to my present self assigned retirement task of tracking cats in a house, looking at wildly varying RSSI signals and wondering what marketeers wet dream made people pursue determining indoor location with gear that's spec'd +-6dbm for RSSI. I begin to wonder... why are they using this idiotic trilateration when triangulation should be within their grasp now, even on-the-cheap? So, finally did a search with proper search terms and "Angle-of-Arrival" and Texas Instruments was at top of search for ble fu. Now I'm confident we will be tracking cats, targeting grandma with granny-cams based on BLE-beacon, helping my robots find the leaky valve under the sink and all manner of I)ndoor P)ositioning S)ystem needs on-the-cheap... I just have to wait a while a bit.

In recent days I've become more confident in using cheapie cat-collar borne ble-beacons to determine which cat is in the box as a proximity problem. I've changed my test scenario to have two simulated ble-beacons (1 RPi3, 1 rpi-zero-W) and detecting beacon transmissions on rpi-zero-W. So far the physically closest always has the statistically strongest RSSI (but still wildly varies).

I've also gained confidence (only thru web r&d) that pet-chips _may_ be detected at significant range to be useful in the "who's-in-the-box" problem.

Consequently, will be moving forward on both approaches of ble-beacon AND RFID and may be silent about that subject for a while.

Near term we will be sticking to the knitting by honing the load-cell weighing of cat/litter/cat-poo/cat-pee and working out details of merging weight data and motion video and moving data on periodic basis to central server.

Andyroo
Posts: 3373
Joined: Sat Jun 16, 2018 12:49 am
Location: Lincs U.K.

Re: catPooPee

Tue Apr 23, 2019 2:20 pm

Good luck with the next stage.

If you need to bounce anything off us - just pop a post here and I’ll waffle away :lol:
Need Pi spray - these things are breeding in my house...

monkeyfork
Posts: 62
Joined: Tue Oct 29, 2013 10:14 am
Location: orlando

Re: catPooPee

Wed May 01, 2019 1:15 am

Andyroo, PhatFil I've been occupied with the more glamorous aspects of project catPooPee, that of cleaning non-absorbing safflower seed cat litter and am zeroing in on a process that passes the sniff test.

I've been doing a bit of light research on the bluetooth 5.1 Angle-of-Arrival and Angle-of-Departure. So far, I still haven't found any case studies that capture and publish real data and consider much of what I'm reading to be marketing BS. Andyroo, since you've actually done some indoor asset tracking I'm sure you're interested in the concept. Soooo, if during your web wandering you happen to find what you'd consider to be solid evidence of working systems please leave a trail of popcorn here.

I'd be especially interested in the use of this Nordic line https://www.nordicsemi.com/?sc_itemid=% ... 35CBF%7D when someone maybe sees fit to make it a I)ndoor P)ositioning S)ystem coprocessor to a rpi-zero-W. I'd do that myself but I have extremely import cat excrement to attend to <joke>.

Return to “Advanced users”