Fascinating… honestly it’s good to learn that sometimes other folks (besides me…) end up having problems that just seem beyond working out (although I usually get stumped way earlier in the task!!). I always enjoy these videos and often try some variation! Sorry these devices didn’t live up to their promised function but I really enjoyed seeing the thought process beside designing the system once’s again!!
In the framework of your airplane trilateration image, if all sensors overestimate, then the intersection is a triangle rather than a point. Just an idea, you could calculate the centre of this triangle rather than hardcoding corrections.
As always, thank you so much for these videos ! OK this time it doesn't work as hoped, but you obvioulsy spent a huge amount of time and you give your work at the end, this is absolutely great ! I hope that someone will eventually figure it out because it could be the basis of something great in an escape room. Looking forward to your next videos :) Thanks again !
The distance from the anchor to the "boat" can be estimated either by a time-of-flight (that would affect the round-trip-time infinitesimally - remember 1ns is a 15cm round-trip) or by RSSI. I suspect that the tags use a combination of these two. Depending on the initialization of the ESP32, I can imagine that a few ns are added here and there. Similarly, dielectrics and reflectors (wood, flesh, walls, plants) shall create multi-path reflections. If you put your hand in between the anchor and the boat - how much does the reported distance change? What "band" (UWB) is used? This is a nice concept in terms of its low-effort deployment; I expect some maturity and learnings by various users (like you) to incrementally improve the precision over the coming months. If several anchors could talk to each other (in addition to talking to the boats) this might too add a layer of self-calibration the the holistic system. Cheers!
I worked a bit with UWB and those Decawave Chips. So regarding your questions: They only use Time-Of-Flight. I can tell you, that (at least with the newer Decawave 3000 Chips), things like wood or a hand don´t change the measured distance at all. It is quite astonishing how good this stuff can work. This is the same technic that Apple is using for their new Air Tags for this neat localization feature. I can also tell you, that currently developing stuff for all of this (especially if you want to work with iPhones), is extremly painfull. Everything is under NDA and you get almost no information from Apple, Decawave or NXP (who also produce pre-production Chips with UWB).
Hi, I recently got 4 ESP32 UWB module and 1 ESP32 UWB Pro with display module. I started experimenting with one or two beacons and the one with display as a TAG. And I ran EXACTLY in the same issue of loosing connections at 80cm or less. It drove me crazy. Till I had the idea to drop the tag one and use another module as the tag. Tadaaaa! I have precise measures at 10m+ without problem, even in 3D and even with a light wall in the way. What a difference! So my idea is that you might have a defective module in your setup... just an idea 🙂
@@skyline128 Have you found a solution to this? We're also looking to do a mini map of sorts for our project but I'm trying to get ahead of any issues before starting our project 😅
Hello. Bir buttun ile butona basılması ile nodemcu dan progmem içindeki wav dosyasından (viole. h) ses çalmak istiyorum. Aslında buton ile değil saat ile alarm saati eşit olduğunda yapmak istiyorum. ama wav dosyasındaki sesi progmem içindeki nodemcu da farklı sekmede olan progmemden sesi çalamıyorum. nasıl bir kod yazılmalıdır bilginiz var mı?
Love the video and your descriptive code. Btw at 39:35 you used the closed form solution. You want to use the iterative solution that minimizes L2 error instead.
I'd never heard of these chips. So I'll be looking further into them! My first thought, while watching, was to use more anchors, but you did that. Shame it didn't work. But, crunchy with the smooth. My only other thought would be using a Kalman filter on the distance data. However, I'm not sure if the lag would make it unusable or not. If you make any further progress, please share. I have a robotics lab at work that I'd love to let robots navigate around indoors, on the cheap, without marvelmind systems. You've also triggered my curiosity about using other wireless tranceivers as TOF sensors(?) Great Video :)
I've taken the documents you've shared but I'm having problems. The position of the anchors on the scene changes when I launch the game and my tag doesn't move. I'm having trouble understanding.
Hello thanks for this video. I would like to use this technology to track a robot mower on an 800m2 plot of land. Do the anchors have enough reach for an 800m2 plot?
This was a couple of years ago but wondering how you went with it ? I think your inaccuracy might have something to do with teh default antenna delay. I didnt see anywhere in the video that you calibrated it, so starting twith that might get you some better results
One of the problems is likely that the timers are not accurate enough for sub centimeter tracking. You are basicly sending a signal with a time value and the tag will then calculate the runtime. If the distances between your stations is to small you won't have much to work with. On a room scale it might work a bit better. Also you could calculate position by having the circles overlap and using the center of that as your location. That way it should also jerk less. These devices are for distances of up to 40 meters (or 120m for the other version). At those distances the offset of 50cm is more negligible. I think with a better timer it might work better though still
I am only just beginning (today) to look at these boards, and I have seen minimum distances of 45cm - this sounds very plausible - although why he could not replicate his original results seems like a coding issue.
hi, I finally received the ESP32 UWB chip, I compiled the tags and anchors, but when I run unity, the tags on unity are not moving, can you give me some advice?
Is there a way to determine the distance between 2 of these devices? I would like to work on a project that will use 2 of these devices to determine their distance from each other. something very simple, ie. if 1 foot apart, turn led green, if 5 feet apart, turn led yellow, if 10 feet apart turn led red.
Excellent intro to these devices. Thank you. BTW... at 3:15 you state that GPS satellites are "geostationary" aka. geosynchronous. They are not. GPS satellites orbit and are always in motion.
Hello! At most, how many TAGs can you use in these systems? Example, in a system with 20 anchors, can I use just one TAG or can I use more than one TAG simultaneously?
It certainly is challenging but it can be done. It's even worse with the DW3000 because the driver provided by Qorvo is closed source. You can certainly get much better range than you experienced with proper configuration, at least up to 100 metres line of sight. But you do have to accept limited accuracy because the antennae on those modules have a significant directional bias, meaning the range measurement depends on the relative orientation of the devices.
Hello. I am trying to communicate two ultra wideband bu01 dw1000 chip using esp32 d1 mini without any breakout board. What i did was to etch a pcb board and only solder the necessary pins to communicate 2 uwb with each other. The pins soldered was the SPI pins (CLK, MOSI, MISO), CS pin ,IRQ pin, RSTn, VSS, and VDD3v3. But it just doesn't work. Any ideas why?
Hi Alistair g8 in depth trial thanks. I'm working on a similar concept trying to track users into Unity. Just wondered if you've tried other Bluetooth beacons or simply GPS in a similar way with Unity? and if so how they compared to these? - tryign to fig out whether they are worth meddling with. 85cm dropout range seemed pretty limiting too
An actual company says that the max precision is 50cm. quote : the location of UWB signals can be determined with an accuracy of less than 50 centimeters (with optimal conditions and deployment), and extremely low latency. (inpixion company home page)
Maybe it is stupid and wrong, but have you checked that SPI config is valid for ESP32? I don't have esp32 but I was playing with DWM1000 few years ago and it was an issue on some boards.
Working in UWB. 11:00 Antenna delay is likely the calibration you are looking for... Maybe. Anyway thé DW chip itself provides a readout of thé ToF in RCTU, so Time unit not (centi)meters. So it looks also likely that the software used is making a wrong assumption about how to concert tof to distance (?)
Thankyou for the insight! Yes, I'm sure that the device *should* be capable of much more and I suspect you're correct that the software isn't behaving correctly, but I just don't have the resources to put in to wading through all that datasheet for what is an incredibly edge use-case (tracking of objects in an escape room!)
Hello I had a question about this project... actually we have been working on a similar project where we are using dw1000 for calculating coordinates..the thing is we want the tag to be as cost effective as possible and easy to manufacture...so is there any way I could send the ranging data from the anchors to the server instead of the tag...we want the tag to just act as a ranging device so that the power consumption and the cost could be low. Thanks in advance ... really liked your projects and it helped me a lot!!
Yes, sure - I was actually uncertain myself what the best way to design the infrastructure was in terms of whether the calculations should be done client-side (and, if so, on the anchor or on the tag?) or server-side. I think the correct approach would probably depend a lot on how many tags/anchors you were planning implementing, and also on what you were then going to do with that information. But you could certainly just get each device to report the ranging time and do all the subsequent processing remotely.
The source of error is the delay of the antenna itself. The antenna delay can be grater or lower depending on the antenna design but the value is fixed for every antenna. The error should be the same no matter the sensing distance. The chip itself doesnt know the antenna thats attached to it so it cant compensate for it. Cheers!
I don't know how the library code works, but I can imagine that mixing up interrupt priorities can really mess with the accuracy of your measurements (?)
Have you tried greater distances, like 10 - 500 metres or more? I suspect it is the same situation as survey. All distances will have a small error that is more noticeable at short ranges, less so at greater distances. I suspect that someone looking for an item that is big enough to house the arduino tag, they should be able to see it at 1/2 metre.
Hi, did the following code from the original Arduino library have any affect? I think in your version, this was removed. //Enable the filter to smooth the distance //DW1000Ranging.useRangeFilter(true); How about a rolling average, or other filtering technique? Do you fancy testing the DWM3000 modules? You might be able to use the same demo boards and swap the modules (hot air gun etc) and run the same experiment.
I can ask you a question please, if I want the same system but without tag, i.e. using the doppler effect of a radar system to monotor a person, is it possible with this tool . I want a sensor that is not worn by a person. thank.
During my test, the data was unstable and always shaking, do you know how to smooth and filter the range? just like the stable performance you showed in the video. Thank you!
It is a sincere review and will be helpful to a lot of people. DO you have experience with any other UWB modules. I am interested in known the maximum range I can expect from these decawave UWB modules in outdoor condition. Thank you.
With clear line of sight and proper configuration they can comfortably communicate easily up to 50 m, and up to 100 m if you're willing to accept some amount of loss. I have no idea what people are doing to get only 10-15 metres.
Your code is almost identical to the example code given by the MakersFab - did the manufacturer of the board steal your code and claimed it as their own?
The schematics are open-source so you can build your own (although I'm not sure I'd bother!): github.com/Makerfabs/Makerfabs-ESP32-UWB/tree/main/hardware
Besides when adding eg. Gps, gprs call, the loop delay adds up and makes distance even wrong. This is impossible to calibrate even for every pair of anchor and tag in a large scale. Further, the makepython A9G shield sim sometimes disconnect without any reason after running for a while. What a waste of time.
@@PlayfulTechnology Finally I understand the behavior. Besides the noise on ranging, there is a delay within the loop code that should be properly tuned.
@@chantana.c Can you tell me how was the performance of after your tuning? Did you have any more methods to improve the performance more after that? I'm planning to use these modules in my project but not sure if the quality is worth it (since the price of these makerfabs modules is so expensive). Thank you!
@@binhnguyenquoc3249 It was better when using wifi. But when using 4G/LTE is is slow. Cannot do any better. The cycle time is limited to the bound of 4G/LTE, if you make it too fast e.g. you will not find the anchor. This will be worse when you have many anchors to scan. I won' recommend for real product but for a toy project, it is fun. 🙂
I'd honestly wait until Adafruit or someone like that takes a crack at it. The in depth complicated documentation needs to be concentrated down and made much simpler and some decent example libraries need to be made. If the manufacturer can not help you (as seen in their sales comments), then it is likely that there might also be mistakes in their hardware design as well. At the very least, customer support seems to be nill.
Don't know how these boards have been implemented but there are a number of 'fundamentals'. The tags and anchors need to be in as much free space as possible, not on a desk or floor and nowhere near metal or wood. Suggest suspending them from a ceiling (nowhere near a room corner). A human touching or coming very close to the boards will cause significant transmission / reception variations. Recommend putting the tag on a tripod (with the antenna well away from the metal tripod) for initial distance testing. Turn the WiFi completely off if possible, take the 'position' output serially from the anchors. The antenna looks odd (planar?), try rotating the anchors and tag and see if the 'distance' changes. In many UWB deployments anchors communicate with each other for timing synchronisation, don't know if this is the case in these boards.
Hi that's an awesome channel, marvelmind robotics does an ultrasound beacon with 20 mm precision however it's about 500 dollars for the minimum 35m range version for all the beacons. perhaps someone will get a germanium 1thz receptor in there and it will be 5mm precise that would be cool. sounds like a new company needs to change these boards ultrasound swarm beacons are not sold open source afaik for rpi that's a dilemna cos i want some for a map robot.
@46:00 I was watching other ESP32 videos and they said that power cables to the ESP32 were a problem. th-cam.com/video/xPlN_Tk3VLQ/w-d-xo.html... the other problem was out of sinc serial communication th-cam.com/video/xPlN_Tk3VLQ/w-d-xo.html ... but every ESP32 is different.
Hi Sir this is the best informative, distinctive ever made on real time location system, Sir I am very excited to do project like this , but idk what would i need thats intimidating me , sir i searched you email so that i should ask but in the bio its not , plz sir reply me so that i should get more information regarding this project.
![[Device Overview] Ultra-wideband Transceiver Module RYUW122](http://i.ytimg.com/vi/HwBakNIPAz4/mqdefault.jpg)
![[Device Overview] Ultra-wideband Transceiver Module RYUW122](/img/tr.png)
![[안방1열 직캠4K] 베이비몬스터 치키타 'DRIP' (BABYMONSTER CHIQUITA FanCam) @SBS Inkigayo 241110](http://i.ytimg.com/vi/tnlxo79cAqY/mqdefault.jpg)
Fascinating… honestly it’s good to learn that sometimes other folks (besides me…) end up having problems that just seem beyond working out (although I usually get stumped way earlier in the task!!). I always enjoy these videos and often try some variation! Sorry these devices didn’t live up to their promised function but I really enjoyed seeing the thought process beside designing the system once’s again!!
In the framework of your airplane trilateration image, if all sensors overestimate, then the intersection is a triangle rather than a point. Just an idea, you could calculate the centre of this triangle rather than hardcoding corrections.
As always, thank you so much for these videos ! OK this time it doesn't work as hoped, but you obvioulsy spent a huge amount of time and you give your work at the end, this is absolutely great ! I hope that someone will eventually figure it out because it could be the basis of something great in an escape room. Looking forward to your next videos :) Thanks again !
Bluetooth Low Energy it is. Anchors in the wild of cities just wont cut it. Great vid thank you.
The distance from the anchor to the "boat" can be estimated either by a time-of-flight (that would affect the round-trip-time infinitesimally - remember 1ns is a 15cm round-trip) or by RSSI. I suspect that the tags use a combination of these two. Depending on the initialization of the ESP32, I can imagine that a few ns are added here and there. Similarly, dielectrics and reflectors (wood, flesh, walls, plants) shall create multi-path reflections. If you put your hand in between the anchor and the boat - how much does the reported distance change? What "band" (UWB) is used?
This is a nice concept in terms of its low-effort deployment; I expect some maturity and learnings by various users (like you) to incrementally improve the precision over the coming months. If several anchors could talk to each other (in addition to talking to the boats) this might too add a layer of self-calibration the the holistic system.
Cheers!
I worked a bit with UWB and those Decawave Chips. So regarding your questions:
They only use Time-Of-Flight.
I can tell you, that (at least with the newer Decawave 3000 Chips), things like wood or a hand don´t change the measured distance at all. It is quite astonishing how good this stuff can work.
This is the same technic that Apple is using for their new Air Tags for this neat localization feature.
I can also tell you, that currently developing stuff for all of this (especially if you want to work with iPhones), is extremly painfull. Everything is under NDA and you get almost no information from Apple, Decawave or NXP (who also produce pre-production Chips with UWB).
Hi. Naveen here. Thanks for showing this. Nice talking with you on facebook.
Can you do this with BLE instead of UWB? The dw1000 boards are kinda expensive…
love the content. looking forward to all the videos you make in 2022!
Hi, I recently got 4 ESP32 UWB module and 1 ESP32 UWB Pro with display module. I started experimenting with one or two beacons and the one with display as a TAG. And I ran EXACTLY in the same issue of loosing connections at 80cm or less. It drove me crazy. Till I had the idea to drop the tag one and use another module as the tag. Tadaaaa! I have precise measures at 10m+ without problem, even in 3D and even with a light wall in the way. What a difference! So my idea is that you might have a defective module in your setup... just an idea 🙂
Any advice on this? I'm looking to create a class project using UWB modules with unity for a mobile game
@@skyline128 Have you found a solution to this? We're also looking to do a mini map of sorts for our project but I'm trying to get ahead of any issues before starting our project 😅
Hello. Bir buttun ile butona basılması ile nodemcu dan progmem içindeki wav dosyasından (viole. h) ses çalmak istiyorum. Aslında buton ile değil saat ile alarm saati eşit olduğunda yapmak istiyorum. ama wav dosyasındaki sesi progmem içindeki nodemcu da farklı sekmede olan progmemden sesi çalamıyorum. nasıl bir kod yazılmalıdır bilginiz var mı?
Love the video and your descriptive code. Btw at 39:35 you used the closed form solution. You want to use the iterative solution that minimizes L2 error instead.
Excellent video, very insightful! Keep it up
I'd never heard of these chips. So I'll be looking further into them! My first thought, while watching, was to use more anchors, but you did that. Shame it didn't work. But, crunchy with the smooth. My only other thought would be using a Kalman filter on the distance data. However, I'm not sure if the lag would make it unusable or not.
If you make any further progress, please share. I have a robotics lab at work that I'd love to let robots navigate around indoors, on the cheap, without marvelmind systems. You've also triggered my curiosity about using other wireless tranceivers as TOF sensors(?) Great Video :)
Worked on this for 3d positionning robot, it's very not robuste, could'nt achieve more than 30/40cm in 3d accuracy in bests conditions
I've taken the documents you've shared but I'm having problems. The position of the anchors on the scene changes when I launch the game and my tag doesn't move. I'm having trouble understanding.
Donde se comora el odulo y cuanto cuesta?
Hello thanks for this video. I would like to use this technology to track a robot mower on an 800m2 plot of land. Do the anchors have enough reach for an 800m2 plot?
You should try the DW3000 modules on higher frequency and apply filters. Apparently range is lots of meters.
This was a couple of years ago but wondering how you went with it ? I think your inaccuracy might have something to do with teh default antenna delay. I didnt see anywhere in the video that you calibrated it, so starting twith that might get you some better results
+1 I would like to know what happened with all of these. I saw (and would like to buy them) DWM3000 chip, maybe this chip will be better. 🧐
One of the problems is likely that the timers are not accurate enough for sub centimeter tracking. You are basicly sending a signal with a time value and the tag will then calculate the runtime. If the distances between your stations is to small you won't have much to work with. On a room scale it might work a bit better. Also you could calculate position by having the circles overlap and using the center of that as your location. That way it should also jerk less. These devices are for distances of up to 40 meters (or 120m for the other version). At those distances the offset of 50cm is more negligible. I think with a better timer it might work better though still
I'm almost sure that's exactly the issue here, I experimented a bit with dwm1001 and got similiar results at close ranges
I am only just beginning (today) to look at these boards, and I have seen minimum distances of 45cm - this sounds very plausible - although why he could not replicate his original results seems like a coding issue.
hi, I finally received the ESP32 UWB chip, I compiled the tags and anchors, but when I run unity, the tags on unity are not moving, can you give me some advice?
Is there a way to determine the distance between 2 of these devices? I would like to work on a project that will use 2 of these devices to determine their distance from each other. something very simple, ie. if 1 foot apart, turn led green, if 5 feet apart, turn led yellow, if 10 feet apart turn led red.
Excellent intro to these devices. Thank you. BTW... at 3:15 you state that GPS satellites are "geostationary" aka. geosynchronous. They are not. GPS satellites orbit and are always in motion.
Hello! At most, how many TAGs can you use in these systems? Example, in a system with 20 anchors, can I use just one TAG or can I use more than one TAG simultaneously?
It certainly is challenging but it can be done. It's even worse with the DW3000 because the driver provided by Qorvo is closed source. You can certainly get much better range than you experienced with proper configuration, at least up to 100 metres line of sight. But you do have to accept limited accuracy because the antennae on those modules have a significant directional bias, meaning the range measurement depends on the relative orientation of the devices.
Could you do a video of what to use or improve in case of disasters like earthquake locating uwb or Lorawan thank you.
Thanks for the honesty it's still a great video :)
Thank you, very high quality tutorial.
Where do I find the schematic to make my own board? I can't seem to find it on the GitHub link provided?
Would it be precise enough to do human body motion with tracking of several probes on the suit?
No. If you want to so skeletal tracking I recommend picking up a Kinect sensor - you can get the 1st gen. ones pretty cheap and they're easy to hack.
Hello. I am trying to communicate two ultra wideband bu01 dw1000 chip using esp32 d1 mini without any breakout board. What i did was to etch a pcb board and only solder the necessary pins to communicate 2 uwb with each other. The pins soldered was the SPI pins (CLK, MOSI, MISO), CS pin ,IRQ pin, RSTn, VSS, and VDD3v3. But it just doesn't work. Any ideas why?
Separate CS for both? Check power etc.
Hi
How to implement it outdoors, where anchors are not in the same location ?
Hi Alistair g8 in depth trial thanks. I'm working on a similar concept trying to track users into Unity. Just wondered if you've tried other Bluetooth beacons or simply GPS in a similar way with Unity? and if so how they compared to these? - tryign to fig out whether they are worth meddling with. 85cm dropout range seemed pretty limiting too
An actual company says that the max precision is 50cm. quote : the location of UWB signals can be determined with an accuracy of less than 50 centimeters (with optimal conditions and deployment), and extremely low latency. (inpixion company home page)
Thanks for reporting your discoveries! Subscribing
Hi, I'd like to measure breathing using UWB radar. Is this possible? Thanks
Chest mounted gyroscope would be better
Maybe it is stupid and wrong, but have you checked that SPI config is valid for ESP32? I don't have esp32 but I was playing with DWM1000 few years ago and it was an issue on some boards.
Working in UWB.
11:00 Antenna delay is likely the calibration you are looking for... Maybe.
Anyway thé DW chip itself provides a readout of thé ToF in RCTU, so Time unit not (centi)meters.
So it looks also likely that the software used is making a wrong assumption about how to concert tof to distance (?)
Thankyou for the insight! Yes, I'm sure that the device *should* be capable of much more and I suspect you're correct that the software isn't behaving correctly, but I just don't have the resources to put in to wading through all that datasheet for what is an incredibly edge use-case (tracking of objects in an escape room!)
Thank for sharing
No problem
Nice work sir…
What if the tag and each anchor is combining their distance evaluation and that’s why the result seems to be doubled?
Amazing video!
Please could you explain how calibrate the measure between anchors and tag please?
Anyone know if the ESP32 UBW software ecosystem has gotten any better to the point it's simple to implement for an individual hobbyist?
Hello I had a question about this project... actually we have been working on a similar project where we are using dw1000 for calculating coordinates..the thing is we want the tag to be as cost effective as possible and easy to manufacture...so is there any way I could send the ranging data from the anchors to the server instead of the tag...we want the tag to just act as a ranging device so that the power consumption and the cost could be low. Thanks in advance ... really liked your projects and it helped me a lot!!
Yes, sure - I was actually uncertain myself what the best way to design the infrastructure was in terms of whether the calculations should be done client-side (and, if so, on the anchor or on the tag?) or server-side. I think the correct approach would probably depend a lot on how many tags/anchors you were planning implementing, and also on what you were then going to do with that information. But you could certainly just get each device to report the ranging time and do all the subsequent processing remotely.
The source of error is the delay of the antenna itself. The antenna delay can be grater or lower depending on the antenna design but the value is fixed for every antenna. The error should be the same no matter the sensing distance. The chip itself doesnt know the antenna thats attached to it so it cant compensate for it. Cheers!
I don't know how the library code works, but I can imagine that mixing up interrupt priorities can really mess with the accuracy of your measurements (?)
Have you tried greater distances, like 10 - 500 metres or more? I suspect it is the same situation as survey. All distances will have a small error that is more noticeable at short ranges, less so at greater distances. I suspect that someone looking for an item that is big enough to house the arduino tag, they should be able to see it at 1/2 metre.
wonder if the new dw3000 is any good
Hi, did the following code from the original Arduino library have any affect? I think in your version, this was removed.
//Enable the filter to smooth the distance
//DW1000Ranging.useRangeFilter(true);
How about a rolling average, or other filtering technique?
Do you fancy testing the DWM3000 modules? You might be able to use the same demo boards and swap the modules (hot air gun etc) and run the same experiment.
Beware the register map is not the same
I need this tracker for cam tracking
I can ask you a question please, if I want the same system but without tag, i.e. using the doppler effect of a radar system to monotor a person, is it possible with this tool . I want a sensor that is not worn by a person. thank.
Theorically feasible.
You need to sens a pulse and switch back immediately to RX to listen the echo.
Just a little point GPS Satellites are NOT in geostationary orbit.
During my test, the data was unstable and always shaking, do you know how to smooth and filter the range? just like the stable performance you showed in the video. Thank you!
You can implement a simple rolling average to smooth the incoming values (it won't help fix the other problems described though!)
Can it be used long range
Best content ❤️❤️
It is a sincere review and will be helpful to a lot of people. DO you have experience with any other UWB modules. I am interested in known the maximum range I can expect from these decawave UWB modules in outdoor condition. Thank you.
From my experience, you can expect around 10-15m (at least with the newer Decawave 3000 Chips)
@@michaelneubauer4401 thanks
I've done around 10 meters with dwm1000 at low refresh rate and in Line of sight. Very Bad regarding decawave documentation
@@michaelneubauer4401 sir I have the same issue can you response.,? So I should further explain
With clear line of sight and proper configuration they can comfortably communicate easily up to 50 m, and up to 100 m if you're willing to accept some amount of loss. I have no idea what people are doing to get only 10-15 metres.
Your code is almost identical to the example code given by the MakersFab - did the manufacturer of the board steal your code and claimed it as their own?
can i just using 2 anchors and 1 tag to complete this trilateration?
I'd say yes.
Is there any cheaper alternatives from this one? Makerfabs' esp32 is too expensive.
The schematics are open-source so you can build your own (although I'm not sure I'd bother!): github.com/Makerfabs/Makerfabs-ESP32-UWB/tree/main/hardware
@@PlayfulTechnology thank you
My guess is that the RF antenna and associated tracks are not quite correct.
Very informative video, thank you! Is there a reason to not use a simple esp32 with its onboard wifi for the same purpose?
GNSS/GPS satellites are NOT in geostationary orbits, but their ephemeris is known. Thus, you can calculate their positions.
Besides when adding eg. Gps, gprs call, the loop delay adds up and makes distance even wrong. This is impossible to calibrate even for every pair of anchor and tag in a large scale.
Further, the makepython A9G shield sim sometimes disconnect without any reason after running for a while. What a waste of time.
Can this be used for multi person positioning?
Yes, you can have multiple anchors and multiple tags.
@@PlayfulTechnology Thanks for your answer, I will continue to follow your videos.^^
You should make a cash register puzzle
I have the same problem too. Every run the offset change...
Yeah, it's very frustrating! :(
@@PlayfulTechnology Finally I understand the behavior. Besides the noise on ranging, there is a delay within the loop code that should be properly tuned.
@@chantana.c Can you tell me how was the performance of after your tuning? Did you have any more methods to improve the performance more after that? I'm planning to use these modules in my project but not sure if the quality is worth it (since the price of these makerfabs modules is so expensive). Thank you!
@@binhnguyenquoc3249 It was better when using wifi. But when using 4G/LTE is is slow. Cannot do any better. The cycle time is limited to the bound of 4G/LTE, if you make it too fast e.g. you will not find the anchor. This will be worse when you have many anchors to scan. I won' recommend for real product but for a toy project, it is fun. 🙂
I'd honestly wait until Adafruit or someone like that takes a crack at it. The in depth complicated documentation needs to be concentrated down and made much simpler and some decent example libraries need to be made. If the manufacturer can not help you (as seen in their sales comments), then it is likely that there might also be mistakes in their hardware design as well. At the very least, customer support seems to be nill.
Neal Fords
Great
tutorial please
Don't know how these boards have been implemented but there are a number of 'fundamentals'.
The tags and anchors need to be in as much free space as possible, not on a desk or floor and nowhere near metal or wood. Suggest suspending them from a ceiling (nowhere near a room corner). A human touching or coming very close to the boards will cause significant transmission / reception variations. Recommend putting the tag on a tripod (with the antenna well away from the metal tripod) for initial distance testing.
Turn the WiFi completely off if possible, take the 'position' output serially from the anchors.
The antenna looks odd (planar?), try rotating the anchors and tag and see if the 'distance' changes.
In many UWB deployments anchors communicate with each other for timing synchronisation, don't know if this is the case in these boards.
Usually two way ranging (Timed but without synchro) is used with these board
Hi that's an awesome channel, marvelmind robotics does an ultrasound beacon with 20 mm precision however it's about 500 dollars for the minimum 35m range version for all the beacons. perhaps someone will get a germanium 1thz receptor in there and it will be 5mm precise that would be cool. sounds like a new company needs to change these boards ultrasound swarm beacons are not sold open source afaik for rpi that's a dilemna cos i want some for a map robot.
Time to replace sony’s HORRIBLY OVERPRICED XR body trackers for XR apps! Thanks!
@46:00 I was watching other ESP32 videos and they said that power cables to the ESP32 were a problem. th-cam.com/video/xPlN_Tk3VLQ/w-d-xo.html... the other problem was out of sinc serial communication th-cam.com/video/xPlN_Tk3VLQ/w-d-xo.html ... but every ESP32 is different.
Whitney Plains
Melody Falls
Kaycee Roads
Jocelyn Squares
You actually need 4 anchors if you want a true 3D positioning.
Alvera Radial
Harber Isle
Conroy Shore
Erna Squares
Adriana Skyway
Swaniawski Grove
مرحبا كيف انت
Bauch Field
Thomas Brenda Garcia Melissa Taylor Sarah
Colby Drives
Halvorson Causeway
Ashlynn Mews
0030 Rachelle Field
Angie Plaza
Beahan Harbor
Hi Sir this is the best informative, distinctive ever made on real time location system, Sir I am very excited to do project like this , but idk what would i need thats intimidating me , sir i searched you email so that i should ask but in the bio its not , plz sir reply me so that i should get more information regarding this project.
Mayert Junction
Ernser Knoll
Davis Carol White Anthony Brown William
Wuckert Cape
Sarah Center
Tiana Courts