Impressive work, I was thinking of using this technology to trilaterate 2 tags with 3 anchors and prevent a certain action from being taken if the two tags were close together. In my application, I would need a resolution of at least 1cm, but I noticed in the video the accuracy is 10cm. Is it possible to increase this resolution with more anchors and a correction code? Another point, can we connect all the anchors to a single central controller? And use a small, inexpensive, low-power controller on the tags?
My idea for a correction code would be to initially use 3 anchors and take a reading, storing position 1 composed of 3 distances. Then, take another reading with 4 anchors and create a function that selects from these 4 distances the position closest possible to position 1 and form position 2, thus eliminating the measurement from one of the anchors. Continue this process successively. The advantage is that for applications where the calculation update is much faster than the movement speed of the tags, the natural fluctuations of one distance would not impact trilateration abruptly, only the natural fluctuation of two or more at the same time.
Dear Felipe @felyp3able *Thank you very much* for your interest in our channel and your encouragement. Glad to know you liked our content. *_Q_* - _but I noticed in the video the accuracy is 10cm. Is it possible to increase this resolution with more anchors and a correction code? Another point, can we connect all the anchors to a single central controller? And use a small, inexpensive, low-power controller on the tags?_ ➡️ Yes, you are absolutely right!! As per datasheet of this UWB module mentions, the location accuracy is up to 10cm. We've also tested it, and found that below 10cm distance readings fluctuates drastically. We haven't yet tested this module for more precise localization with multiple Anchors and correction code. So we can not comment on its performance with this approach. ➡️ Although, use of small inexpensive, low-power MCU on tags can be advisable. Since we usually do not require that much processing power on Tag nodes. The codes we use in our in the DIY projects and tutorials are only for educational and learning purposes, they usually are not optimized for critical commercial/Industrial applications. Even for this tutorial/ review DIY project, we have used simple and minimal coding paradigm. We are not checking AT command responses and Error codes of commands. Documentation about this module is very rare and there are also no any pre-compiled libraries available. We have to program them with barematal way. Explore it if you are interested, and let us know about your experience. If you have any queries/issues/suggestions about this Tutorial, or any of our DIY project, feel free to ping us on WhatsApp/Telegram (Links given below) for further support. Tech support - (𝗖𝗵𝗮𝘁 𝗢𝗻𝗹𝘆) Telegram/WhatsApp - 11:00 AM - 05:00 PM (Mon-Fri) IST 🚀 𝗧𝗲𝗹𝗲𝗴𝗿𝗮𝗺 🔗 t.me/make2explore 💬 𝗪𝗵𝗮𝘁𝘀𝗔𝗽𝗽 🔗 bit.ly/3VXGGEe 📩 info@make2explore.com / support@make2explore.com As per time permits we will try our best to help you out. Best Regards
Dear Ritchie @ritchiec5438 *Thank you very much* for your interest in our channel and your encouragement. Glad to know you liked our content. If you have any queries/issues/suggestions about this Tutorial, or any of our DIY project, feel free to ping us on WhatsApp/Telegram (Links given below) for further support. Tech support - (𝗖𝗵𝗮𝘁 𝗢𝗻𝗹𝘆) Telegram/WhatsApp - 11:00 AM - 05:00 PM (Mon-Fri) IST 🚀 𝗧𝗲𝗹𝗲𝗴𝗿𝗮𝗺 🔗 t.me/make2explore 💬 𝗪𝗵𝗮𝘁𝘀𝗔𝗽𝗽 🔗 bit.ly/3VXGGEe 📩 info@make2explore.com / support@make2explore.com As per time permits we will try our best to help you out. Best Regards
Dear Huyền vi @vusonhuyenphuthuat *Thank you very much* for your interest in our channel and your encouragement. Glad to know you liked our content. If you have any queries/issues/suggestions about this Tutorial, or any of our DIY project, feel free to ping us on WhatsApp/Telegram (Links given below) for further support. Tech support - (𝗖𝗵𝗮𝘁 𝗢𝗻𝗹𝘆) Telegram/WhatsApp - 11:00 AM - 05:00 PM (Mon-Fri) IST 🚀 𝗧𝗲𝗹𝗲𝗴𝗿𝗮𝗺 🔗 t.me/make2explore 💬 𝗪𝗵𝗮𝘁𝘀𝗔𝗽𝗽 🔗 bit.ly/3VXGGEe 📩 info@make2explore.com / support@make2explore.com As per time permits we will try our best to help you out. Best Regards
Dear Chandrashekar @ChandrashekarCN *Thank you very much* for your interest in our channel. Glad to know that you liked our content. If you have any queries/issues/suggestions about this Tutorial, or any of our DIY project, feel free to ping us on WhatsApp/Telegram (Links given below) for further support. Tech support - (𝗖𝗵𝗮𝘁 𝗢𝗻𝗹𝘆) Telegram/WhatsApp - 11:00 AM - 05:00 PM (Mon-Fri) IST 🚀 𝗧𝗲𝗹𝗲𝗴𝗿𝗮𝗺 🔗 t.me/make2explore 💬 𝗪𝗵𝗮𝘁𝘀𝗔𝗽𝗽 🔗 bit.ly/3VXGGEe 📩 info@make2explore.com / support@make2explore.com As per time permits we will try our best to help you out. Best Regards
This is an excellent video demonstrating how to measure distance between two RYUW122 modules. It would be awesome to see a similar video showing how to measure two dimensional position using one tag and three anchors.
Dear Craig Austin @aguanut *Thank you very much* for your interest in our channel and your encouragement. Glad to know that you liked our content. *_Q_* - _It would be awesome to see a similar video showing how to measure two dimensional position using one tag and three anchors._ ➡️ Designing *_Trilateration positioning_* using this RYUW122 can be a bit challenging task, as we have to follow baremetal programming approach, since there is no any prebuilt library available for these UWB modules. But it's definitely worth a try. 📌 *Thank you very much* for your invaluable suggestion, We'll definitely try to add that tutorial/experiment in our playlist very soon. Beside If you still have any queries/issues/suggestions about this Tutorial, or any of our DIY project, feel free to ping us on WhatsApp/Telegram (Links given below) for further support. Tech support - (𝗖𝗵𝗮𝘁 𝗢𝗻𝗹𝘆) Telegram/WhatsApp - 11:00 AM - 05:00 PM (Mon-Fri) IST 🚀 𝗧𝗲𝗹𝗲𝗴𝗿𝗮𝗺 🔗 t.me/make2explore 💬 𝗪𝗵𝗮𝘁𝘀𝗔𝗽𝗽 🔗 bit.ly/3VXGGEe 📩 info@make2explore.com / support@make2explore.com Best Regards
@@aguanut I did a machine learning experiment for this a few years ago to simulate a radio network, and I applied gaussian noise to the distance measurements. I used a simple gradient descent method after calculating the mean squared error function and partial derivatives for various distance metrics. I'll try to post a link to some results, but youtube might censor it. The biggest problem I encountered was trying to trilaterate in more than 2 dimensions, which I wanted to accomplish to accommodate multi-level buildings.
@@aguanut Yeah, the links to the graphs were censored. Oh well. I'm going to see if I can adapt the maths to make it work with Newton's method or Chebyshev iteration, but I think that trilateration in 2 dimensions might be practical by taking distance measurements from 3 stationary reference radios, then find the intersections of two circles using the closest two measurements, then taking the intersection point that minimizes the error wrt the third point using Euclidean distance calculated from coordinates and the distance measurement. Btw, distance and latency are mathematically equivalent because latency = distance * (2/c), i.e. latency is just distance scaled by a constant, and distance is latency scaled by the inverse of that constant.
@k98killer You're on the right track! Adapting the math to use Newton's method or Chebyshev iteration could lead to more efficient and accurate solutions. Your approach to trilateration in 2D is spot on: - Take distance measurements from 3 stationary reference radios. - Find the intersections of two circles using the closest two measurements. - Select the intersection point that minimizes the error with respect to the third point using Euclidean distance. This method is known as *"circle-circle intersection"* or we can say *"trilateration with three anchors."* It's a robust and efficient technique for 2D localization. Your observation about distance and latency being mathematically equivalent is also correct. Since latency is directly proportional to distance, you can use either metric interchangeably in your calculations. Some additional things we can consider: --> Use a robust circle-circle intersection algorithm to handle edge cases and numerical stability. --> Consider using a weighted least squares approach to minimize the error, giving more importance to the closest reference points. - -> If you have more than three reference points, you can use a more robust method like multilateration or a Kalman filter to improve accuracy. Keep us updated on your progress! We're excited to see how your project evolves. 💗 Best Regards
Impressive work, I was thinking of using this technology to trilaterate 2 tags with 3 anchors and prevent a certain action from being taken if the two tags were close together. In my application, I would need a resolution of at least 1cm, but I noticed in the video the accuracy is 10cm. Is it possible to increase this resolution with more anchors and a correction code? Another point, can we connect all the anchors to a single central controller? And use a small, inexpensive, low-power controller on the tags?
My idea for a correction code would be to initially use 3 anchors and take a reading, storing position 1 composed of 3 distances. Then, take another reading with 4 anchors and create a function that selects from these 4 distances the position closest possible to position 1 and form position 2, thus eliminating the measurement from one of the anchors. Continue this process successively. The advantage is that for applications where the calculation update is much faster than the movement speed of the tags, the natural fluctuations of one distance would not impact trilateration abruptly, only the natural fluctuation of two or more at the same time.
Dear Felipe @felyp3able
*Thank you very much* for your interest in our channel and your encouragement. Glad to know you liked our content.
*_Q_* - _but I noticed in the video the accuracy is 10cm. Is it possible to increase this resolution with more anchors and a correction code? Another point, can we connect all the anchors to a single central controller? And use a small, inexpensive, low-power controller on the tags?_
➡️ Yes, you are absolutely right!! As per datasheet of this UWB module mentions, the location accuracy is up to 10cm. We've also tested it, and found that below 10cm distance readings fluctuates drastically. We haven't yet tested this module for more precise localization with multiple Anchors and correction code. So we can not comment on its performance with this approach.
➡️ Although, use of small inexpensive, low-power MCU on tags can be advisable. Since we usually do not require that much processing power on Tag nodes.
The codes we use in our in the DIY projects and tutorials are only for educational and learning purposes, they usually are not optimized for critical commercial/Industrial applications. Even for this tutorial/ review DIY project, we have used simple and minimal coding paradigm. We are not checking AT command responses and Error codes of commands.
Documentation about this module is very rare and there are also no any pre-compiled libraries available. We have to program them with barematal way. Explore it if you are interested, and let us know about your experience.
If you have any queries/issues/suggestions about this Tutorial, or any of our DIY project, feel free to ping us on WhatsApp/Telegram (Links given below) for further support.
Tech support - (𝗖𝗵𝗮𝘁 𝗢𝗻𝗹𝘆) Telegram/WhatsApp - 11:00 AM - 05:00 PM (Mon-Fri) IST
🚀 𝗧𝗲𝗹𝗲𝗴𝗿𝗮𝗺 🔗 t.me/make2explore
💬 𝗪𝗵𝗮𝘁𝘀𝗔𝗽𝗽 🔗 bit.ly/3VXGGEe
📩 info@make2explore.com / support@make2explore.com
As per time permits we will try our best to help you out.
Best Regards
Very professional👍👍
Dear Ritchie @ritchiec5438
*Thank you very much* for your interest in our channel and your encouragement. Glad to know you liked our content.
If you have any queries/issues/suggestions about this Tutorial, or any of our DIY project, feel free to ping us on WhatsApp/Telegram (Links given below) for further support.
Tech support - (𝗖𝗵𝗮𝘁 𝗢𝗻𝗹𝘆) Telegram/WhatsApp - 11:00 AM - 05:00 PM (Mon-Fri) IST
🚀 𝗧𝗲𝗹𝗲𝗴𝗿𝗮𝗺 🔗 t.me/make2explore
💬 𝗪𝗵𝗮𝘁𝘀𝗔𝗽𝗽 🔗 bit.ly/3VXGGEe
📩 info@make2explore.com / support@make2explore.com
As per time permits we will try our best to help you out.
Best Regards
nice
Dear Huyền vi @vusonhuyenphuthuat
*Thank you very much* for your interest in our channel and your encouragement. Glad to know you liked our content.
If you have any queries/issues/suggestions about this Tutorial, or any of our DIY project, feel free to ping us on WhatsApp/Telegram (Links given below) for further support.
Tech support - (𝗖𝗵𝗮𝘁 𝗢𝗻𝗹𝘆) Telegram/WhatsApp - 11:00 AM - 05:00 PM (Mon-Fri) IST
🚀 𝗧𝗲𝗹𝗲𝗴𝗿𝗮𝗺 🔗 t.me/make2explore
💬 𝗪𝗵𝗮𝘁𝘀𝗔𝗽𝗽 🔗 bit.ly/3VXGGEe
📩 info@make2explore.com / support@make2explore.com
As per time permits we will try our best to help you out.
Best Regards
💖💖💖💖
Dear Chandrashekar @ChandrashekarCN
*Thank you very much* for your interest in our channel. Glad to know that you liked our content.
If you have any queries/issues/suggestions about this Tutorial, or any of our DIY project, feel free to ping us on WhatsApp/Telegram (Links given below) for further support.
Tech support - (𝗖𝗵𝗮𝘁 𝗢𝗻𝗹𝘆) Telegram/WhatsApp - 11:00 AM - 05:00 PM (Mon-Fri) IST
🚀 𝗧𝗲𝗹𝗲𝗴𝗿𝗮𝗺 🔗 t.me/make2explore
💬 𝗪𝗵𝗮𝘁𝘀𝗔𝗽𝗽 🔗 bit.ly/3VXGGEe
📩 info@make2explore.com / support@make2explore.com
As per time permits we will try our best to help you out.
Best Regards
This is an excellent video demonstrating how to measure distance between two RYUW122 modules. It would be awesome to see a similar video showing how to measure two dimensional position using one tag and three anchors.
Dear Craig Austin @aguanut
*Thank you very much* for your interest in our channel and your encouragement. Glad to know that you liked our content.
*_Q_* - _It would be awesome to see a similar video showing how to measure two dimensional position using one tag and three anchors._
➡️ Designing *_Trilateration positioning_* using this RYUW122 can be a bit challenging task, as we have to follow baremetal programming approach, since there is no any prebuilt library available for these UWB modules. But it's definitely worth a try.
📌 *Thank you very much* for your invaluable suggestion, We'll definitely try to add that tutorial/experiment in our playlist very soon.
Beside If you still have any queries/issues/suggestions about this Tutorial, or any of our DIY project, feel free to ping us on WhatsApp/Telegram (Links given below) for further support.
Tech support - (𝗖𝗵𝗮𝘁 𝗢𝗻𝗹𝘆) Telegram/WhatsApp - 11:00 AM - 05:00 PM (Mon-Fri) IST
🚀 𝗧𝗲𝗹𝗲𝗴𝗿𝗮𝗺 🔗 t.me/make2explore
💬 𝗪𝗵𝗮𝘁𝘀𝗔𝗽𝗽 🔗 bit.ly/3VXGGEe
📩 info@make2explore.com / support@make2explore.com
Best Regards
@@make2explore Awesome, I look forward to more of your videos. Thank You!
@@aguanut I did a machine learning experiment for this a few years ago to simulate a radio network, and I applied gaussian noise to the distance measurements. I used a simple gradient descent method after calculating the mean squared error function and partial derivatives for various distance metrics. I'll try to post a link to some results, but youtube might censor it. The biggest problem I encountered was trying to trilaterate in more than 2 dimensions, which I wanted to accomplish to accommodate multi-level buildings.
@@aguanut Yeah, the links to the graphs were censored. Oh well. I'm going to see if I can adapt the maths to make it work with Newton's method or Chebyshev iteration, but I think that trilateration in 2 dimensions might be practical by taking distance measurements from 3 stationary reference radios, then find the intersections of two circles using the closest two measurements, then taking the intersection point that minimizes the error wrt the third point using Euclidean distance calculated from coordinates and the distance measurement. Btw, distance and latency are mathematically equivalent because latency = distance * (2/c), i.e. latency is just distance scaled by a constant, and distance is latency scaled by the inverse of that constant.
@k98killer You're on the right track! Adapting the math to use Newton's method or Chebyshev iteration could lead to more efficient and accurate solutions.
Your approach to trilateration in 2D is spot on:
- Take distance measurements from 3 stationary reference radios.
- Find the intersections of two circles using the closest two measurements.
- Select the intersection point that minimizes the error with respect to the third point using Euclidean distance.
This method is known as *"circle-circle intersection"* or we can say *"trilateration with three anchors."* It's a robust and efficient technique for 2D localization.
Your observation about distance and latency being mathematically equivalent is also correct. Since latency is directly proportional to distance, you can use either metric interchangeably in your calculations.
Some additional things we can consider:
--> Use a robust circle-circle intersection algorithm to handle edge cases and numerical stability.
--> Consider using a weighted least squares approach to minimize the error, giving more importance to the closest reference points.
- -> If you have more than three reference points, you can use a more robust method like multilateration or a Kalman filter to improve accuracy.
Keep us updated on your progress! We're excited to see how your project evolves. 💗
Best Regards