#203

Fast measurement of distances became a critical technology because it enables, together with Artificial Intelligence, autonomous movements of cars or planes. Today we will get an overview and compare readily available devices for makers.
Links:
Sharp Reflective Sensor: s.click.aliexpress.com/e/b6JbDLFi
TFminiLIDAR: s.click.aliexpress.com/e/cth5G43q
Garmin LIDAR Lite: www.sparkfun.com/products/14032
TFmini Library: github.com/opensensinglab/tfmini
Garmin library: On Sparkfun page
StepperBot: github.com/Brezensalzer/Stepp...
Rangefinder video (German): • Neue Erkenntnisse zum ...
Supporting Material and Blog Page: www.sensorsiot.org
Github: www.github.com/sensorsiot
My Patreon Page: / andreasspiess
My Bitcoin address: 19FSmqbBzb5zsYB1d8Bq4KbxVmezToDNTV
If you want to support the channel, please use the links below to start your shopping. No additional charges for you, but I get a commission (of your purchases the next 24 hours) to buy new stuff for the channel
For Banggood bit.ly/2jAQEf4
For AliExpress: bit.ly/2B0yTLL
For ebay.com: ebay.to/2DuYXBp
profile.php?...
/ spiessa
www.instructables.com/member/...
Please do not try to Email me or invite me on LinkedIn. These communication channels are reserved for my primary job
Equipment in my lab: www.sensorsiot.org/my-lab/

[@AndreasSpiess]

The Sharp uses Triangulation. I did not read the datasheet correctly!

[@ddegn]

The one in your video uses triangulation but there are IR sensors which rely on brightness levels. I'm sure you know this but I thought I'd mention it in case others aren't aware.

[@ddegn]

Mai Mariarti Do you want to use the sensor underwater? If so, you need a different type of ultrasound sensor. If you want to use the sensor above water, then there are several options for waterproof ultrasound sensors. I believe the guy with the Swiss accent has shown us some of these waterproof sensors. Maxbotix makes weather resistant ultrasound sensors. Here's one at SparkFun: www.sparkfun.com/products/11724 If you need help finding a waterproof ultrasound transducer, let us know so we can help you look.

[@zahiddogan]

Hello Buddy , do you know what makes tof lidar receivers detect reflected light over 100 meters ? I cant imagine a tiny beam of light hits target 100 meters away and return a tiny receiver .

[@AndreasSpiess]

Look at the rangefinder video. It worked over 1500 meters

[@whyindeed9937]

Polygonal mirrors can be recovered from old laser printers. They make a superb laser scanners.

[@Aegilops]

"to avoid reading the manuals" ... superb as always Andreas

[@josecampino7476]

Very good and concise explanation of the phase-shift measurement method. Thanks again Andreas !

[@AndreasSpiess]

You are welcome!

[@RP6conrad]

Again a very good explanation of the laser range finders ! One remark : the first sensor in the video (Sharp) is also based on triangulation, and has a psd inside.

[@AndreasSpiess]

You are right! I did not read the datasheet correctly!

[@eddyfontaineyoutu100]

Very interesting summary of the different working principles ! 👍👍👍

[@AndreasSpiess]

Thank you!

[@vfinch8483]

Very nice! Just the right amount of information and a great comparison.

[@AndreasSpiess]

Thank you for your nice words!

[@GeekMustHave]

Thank you for this video. I understand LIDAR much better. Keep broadcasting!!

[@AndreasSpiess]

You are welcome!

[@James-fe7wd]

Amazing information right here, thank you for sharing it!

[@AndreasSpiess]

You are welcome!

[@gabriel_gelli]

Excellent video and explanation! Helped a lot! Thanks!

[@AndreasSpiess]

You are welcome!

[@Herebuss]

Very good summation!

[@AndreasSpiess]

Thank you!

[@MrZetor]

Thanks for yet another valuable video! I wouldn't be too happy if I forked out almost 200eur for a LIDAR module (not even a rotating one) that is specified for a 40 meter range, and then discovered that it cannot reliably measure said range even in a dark garage with zero sunlight, _even_ when the target is a flat, white plate, perpendicular to the beam! What comes to short range LIDARs, I would suggest scavenging used and/or broken Neato or Vorwerk robo-vacs. They (at least the high end models) have featured a LIDAR at least for five years, and from what I have read, it is very good and well documented by the hacker community.

[@AndreasSpiess]

1. Life is not always a success. This is part of the concept of this channel: I test and you decide if you need the stuff (and maybe avoid disappointments). For people with the need of far-reaching obstacle avoidance, the Garmin might be the only possibility.2. Thanks for the hint for another source of Lidars. Another viewer also referenced to the Neato. But I do not know these brands here. So I have to stick to what I have for the moment.

[@joseluisalcaraz9071]

Very instructive! Thanks for sharing.

[@AndreasSpiess]

You are welcome!

[@alcanceweb]

Thank you for this analysis.

[@AndreasSpiess]

You are welcome!

[@zhenmeili8858]

Thanks great comparison video!

[@AndreasSpiess]

You are welcome

[@zhitailiu3876]

Dude, I love your videos! Always guide me the latest tech! Thanks!

[@AndreasSpiess]

You are welcome!

[@roberto660260]

Hi Mr. Spiess, I congratulate you on the detailed videos you offer us. They are educational videos and I appreciate them very much, word of elettronics engineer. Very interesting this last video on the sensors to determine the distance, congratulations. A curiosity from the video number 200, for the follow videos ,you did not use the numbering, very convenient in my opinion, there is an reason. Thank you for your commitment, an idea for her: to create instant books on specific but complete subjects and detailed.

[@AndreasSpiess]

The numbering will appear again, for the moment a few videos behind (200 should be numbered now)

[@PhG1961]

Excellent video and well explained. Thnxs. Btw, I love your garage, so much space...

[@AndreasSpiess]

Because I share it with 100 other families...

[@PhG1961]

Wow, 100 is a lot. I thought it was a smaller community/property.

[@AndreasSpiess]

300 parking lots on two floors. Including a atomic bunker, of course ;-)

[@PhG1961]

Yes I've seen the video about the bunker... impressive. From what I remember, almost every (new build) house has an atomic bunker in Switserland. I don't even have a cellar, which would be nice to keep some wine. That's why I drink my wine rather fast.

[@AndreasSpiess]

Good strategy, I think!

[@richard_wenner]

Beautifully explained and demonstrated but lacked an Episode number!

[@sethrd999]

Totally agree with this comment, what an amazing reference

[@AndreasSpiess]

They already returned in older videos.

[@northshorepx]

Patreon here: Looks like this will be in all cars within 5-10 years for obstacle detection. Your experiments are very interesting +Andreas.

[@AndreasSpiess]

We will see if we still call them "cars" then. Maybe they are more "taxis". I do not think it would be of big value to replace me as a driver if I still have to own the car. My work of driving is free of charge. I think a big difference will be that I just order "transportation" and it will be cheaper than driving my own car because these vehicles can be operated 24/7 and not only one hour or less a day.

[@northshorepx]

I do agree.

[@Electronieks]

northshorepx is already implemented in volvo

[@AndreasSpiess]

Not in mine. I still have to drive on my own :-(

[@GRBtutorials]

Owning cars is cheaper than taxis if you need to make a lot of km a day, like if you need to commute long distances. Even with self-driving cars, the taxi company has to pay for the cars, the fuel (electricity since we're in the future) and the maintenance. The taxi company also has to make a profit. So it isn't likely to be cheap.

[@avejst]

Thanks for sharing 😀👍 Interesting facts 😀

[@AndreasSpiess]

You are welcome!

[@tsbimbra]

Thanks for sharing valuable knowledge. Very interesting

[@AndreasSpiess]

You are welcome!

[@purge98]

@@AndreasSpiess I like your Mexican accent.

[@tetraquark2402]

Cool I will wait a bit and buy one

[@AndreasSpiess]

:-)

[@harsh9558]

Man awesome video!

[@AndreasSpiess]

Thank you!

[@AThreeDogNight]

Good video, thank you.

[@AndreasSpiess]

You are welcome!

[@davidsteinhour5562]

Excellent! I learned a lot. I called those Sharp sensors "ET sensors" because they look a little alien. They worked intermittently for my robotics club but we still used them a ton.

[@AndreasSpiess]

Thanks for sharing your experience!

[@seamusfitzgerald6506]

Andreas Spiess Hello Andreas . I don’t mean to hound you but I could do with a bit of advice please . As you know in our world at present we have to stay a safe distance of 2 meters apart , but that’s not always ease . I want to build a system with possible distance sensors enclosed in a badge or bracelet . Example when person A approaches person B and if person A breaches the 2 meters the unit activates for a shoe RT time either flashes or sounds or even vibrate . Can you please point me in the right direction on where to get the correct parts and what parts I require . Thank you in advance Seamus Fitzgerald

[@NPObousai]

探してた情報が見つかりました。ありがとう

[@AndreasSpiess]

どういたしまして

[@ats89117]

Another fantastic video! I can't help mentioning a trivia point: When using the phase shift from multiple frequencies to determine range, the correct range can easily be determined using an algorithm called the "Chinese Remainder Theorem". It is a very old algorithm and always struck me as almost magical when I was working on navigation systems for Kearfott! :-)

[@AndreasSpiess]

So you must be a specialist (looking at the homepage of the company)! So you used this theorem to determine the different frequencies used?

[@ats89117]

I left Kearfott 32 years ago after dragging them kicking and screaming into the digital signal processing era by using the original TMS320 on my Correlation Velocity Sensor program. Back then, the TMS320 cost about $1,500 which was a ton of money in the early 80s, and all the old analog guys didn't see any need for this new fangled stuff! :-D The current state of the art in CRT (Chinese Remainder Theorem) is detailed in a 2017 paper called "Robustness in Chinese Remainder Theorem" by H. Xiao et. al. Dr. Xiao is a professor at MIT. I would put a link to it here, but apparently copying into the comments isn't allowed. At Kearfott, we used CRT in our navigation equipment wherever we could, for both frequency estimation when we only had a short glimpse at a signal, and also for range estimation based on multiple frequencies. These were both key to our Doppler Navigation Systems since these systems always needed to provide range estimates for both altimetry and obstacle avoidance purposes. The biggest problem back then was as Xiao described in his paper, insufficient SNR would result in phase errors that could lead to the calculating the wrong number of cycles, or what we referred to as gross error, as opposed to a fine error where the integer number of cycles were correct, and the error was limited to the residual. It seems this is much less of a problem now than it used to be, but for whatever reason, it's still not commonly known that there is a reasonable simple algorithm for solving for the number of elapsed cycles based on the phase residual in a multi-frequency system.

[@AndreasSpiess]

Thanks for the hint. I found it. But it will take some time to read and understand it ;-) BTW you can post links in the comments. Many others do it too. Sometimes they end in my spam folder. But if the links are ok I can allow them.

[@electronic7979]

Excellent

[@AndreasSpiess]

:-)

[@Zhaymoor]

new technology to me :O , thank you

[@AndreasSpiess]

You are welcome!

[@climbeverest]

Incredible

[@AndreasSpiess]

:-)

[@airflo1]

Danke fur dein video. Ich mag lernen. ^^

[@AndreasSpiess]

Bitte, gern geschehen!

[@__mk_km__]

How about using time-of-flight together with phase shift detection? You shoot laser modulated with a high-frequency signal, measure the time it was travelling, and then look at its phase-shift for even more accurate distance. They can be the same accuracy as the time-of-travel sensor, while costing less because they don't require precision as high for the Time-to-Digital converter. Also, the drop in speed is insignificant compare to the ToT sensor

[@AndreasSpiess]

I am not a specialist. Maybe somebody does it if it is so much better or cheaper.

[@DreadX10]

Those do exist but you need a (military) security clearance to play with them. So I think it's more a matter of patents that stop this tech from being readily available.

[@HobbiesNTech]

You put in a lot of effort in your videos. The findings are well researched. Love your videos. Please keep up your good work. You earned a subscriber. One thing I need to ask you. Do these lidar sensors work with edible oils. I have a projet where we have to measure oil level in tanks 25 metres deep. Will the TF02 work in such a scenario?

[@AndreasSpiess]

Welcome to the channel! The sensors depend on a reflected signal. I assume a quiet oil surface will reflect the beam. However, I do not know on the maximum distance

[@NimaSajedi]

@Andread Spiess, at 4:40, result of equation is wrongly written 10m. It should be 30 meters.

[@AndreasSpiess]

You are right! Thank you.

[@user-fc2ju4cu3w]

thank you

[@AndreasSpiess]

:-)

[@SuburbanDon]

Until recently I worked for a company that measures the tightness of large bolts by bouncing a 20MHz pulse to the end of the bolt and back. This was called time of flight. A longer time indicated the stretch of the bolt.

[@AndreasSpiess]

Interesting they use the same principle with radio frequencies

[@stefanmayer444]

11:43 How does the sensor calibrate itself? Does it calibrate the measured distance or is it an adjustment of ambient lighting? Very good Video as always! You're such a valuable source of information for us makers! Thanks 😊

[@AndreasSpiess]

I assume it is for the light intensity. Distance calibration would need a known distance, I think.

[@stefanmayer444]

@@AndreasSpiess Thanks :-)

[@moellersworkshop2116]

Great demo, thanks! 1061? Possible photonic cross talk by both devices operating at the same time? Also, the laser in the GARMIN might be FLOODING due to closeness, the device is designed to function at 40m right?

[@AndreasSpiess]

The Garmin has a range of Range: 5 cm to 40 meters. Crosstalk is not likely because of the different timing.

[@udith-pz4tv]

Thanks........

[@AndreasSpiess]

:-)

[@alexdoan6943]

GDay, Andreas, will you ever Vlog about connectivity between devices and IoT platforms (open-source or commercial)? 'How to' videos are on demand. i.e. how to connect sensors to ThingsBoard . Many tnx

[@AndreasSpiess]

Maybe I will do such a video. At least it is on my list...

[@Thomas72B]

YDLIDAR X4 Lidar - 360-Grad-Laser !! This sensor could also be very interesting. You can order it from Amazon. The price is also okay.

[@AndreasSpiess]

It is a similar design as I showed in this video and used in another one. But I did not test it so far.

[@deepsatsangi8507]

Excellent. When are you. Going. To come out with very long distance obstacle detection LIDAR ?

[@AndreasSpiess]

For the moment, I have no such plans. :-(

[@rickharold7884]

Cool

[@AndreasSpiess]

:-)

[@suicidalbanananana]

That 360 lidar you shortly show, is that a neato (a certain brand of robot vacuum) lidar? if yes plz do a video on it, ive been eyeballing those things for years yet im still to scared to buy one, thinking i wont get it to work on a arduino or rbpi

[@AndreasSpiess]

I got it from dfrobots and I plan to do a video with it.

[@suicidalbanananana]

Perfect :)

[@rricd]

Hi, Andreas, first of all, I folow all your tutorial and there are greate, thank you for all the time and dedication that you put in it. I want to ask, how fast is the detection in the sharp sensor, if you put it in the robot car, what would be the maximun speed of the car? And, did you have to pos-process the informacion of the sensor.? Or is done is real time?

[@AndreasSpiess]

This is an old video and I am an old man. So I do not remember the details. I am sure you will find answers if you read the specifications. I assume you have to do post-processing to do decisions based on the sensor data. I only tested the sensors.

[@spreeuferberlin]

nice. thx.

[@AndreasSpiess]

:-)

[@letsshare2314]

Thank you for this amazing video...!! Can you please tell the reflection angel the GARMIN makes ...for tf mini it's 2.3 degree

[@AndreasSpiess]

I do not know. Maybe it is in the data sheet?

[@letsshare2314]

@@AndreasSpiess thank you for replying...checked data sheet it's mentioned as length/100...have to get one and test however .

[@hosseinhayati9820]

Thank you for sharing your knowledge and experience :) In a project Iæve use arduino ultra sound sensors, but they are not good enough. Inside building they are accurate and perfomance is good, but when I try it outside in the dark and snow on the ground the performance is poor! Not accurate reading and not reliable. I need to measure between 20cm and 100cm. A good solution maybe even be a combination of sensor types to handle darkness and differe surfaces, temperatur etc.. What kind of sensor do you recommand?

[@AndreasSpiess]

I have no recommendations for such situations. You have to try which is better.

[@fugit185]

For obstacle avoidance you should understand the difference between laser and sonar. Lasers are always beams and if you use them for obstacle avoidance, the good is that you can easily get the angle where obstacle is but the weakness is that because it is beam, it can go "through" a obstacle where a robot coul still collide. Then sonar, it's coverage is a cone shaped so it is good to find any shape obsatcle but you need make it to rotate to get angle for obstacle. Therefore, in our robot we use both, sonar and lidar for object detection, and lidar for distance information

[@AndreasSpiess]

You are right, they have quite different properties.

[@drjwilber]

stated that no cmos linear sensors - have some cmos dram 4k as well as 64 k . Took the top off the memory chip to 'play' with it . Can send if know where to send - plus data

[@ryanbayona409]

Hi Andreas, for short distance measurement ( within 1.5 meter ) which one do you recommend? TFMini or VL53L1X? Which one is more accurate?

[@AndreasSpiess]

I never did a comparison. Also take into account if it is indoors or outdoors. the VLX... do not work very good outdoors.

[@PHCareli]

which sensor would you use for fast robots with short-range detection?

[@AndreasSpiess]

Your choice. I gave you the basics for decision making.

[@ChristopherBugeja]

If the tiny lidar sensor is used in a well with water, would it need a white surface as a reflector? Or would you reckon that the reflection from the surface of the water be enough?

[@AndreasSpiess]

I do not know. The light has to be reflected back to the Lidar, otherwise, it cannot be detected.

[@TonyLehto]

When can we expect the SDR-videos? 😁

[@AndreasSpiess]

This summer...

[@arunprakash7385]

Dear @Andreas Spiess, thank you, for a nice and informative video. I want to build a distance sensor that should be precise for distances of 2m to 10m. It should have an accuracy of 0.1 % for shorter distances like2-5m and can have an accuracy of 1% at 10m or more. Is this possible in one range finder or I have to use a different finder for different distances. Also which would be better a Laser-based or sound-based for above requirement.

[@AndreasSpiess]

You have to read the different datasheets and search for what you need.

[@arunprakash7385]

@@AndreasSpiess, I have seen the datasheets of lidars available in my country. I didn't come across a LIDAR that has 0.5% accuracy at 5m. also the fact that the specified accuracy is not seen reality. From your experience, please suggest me a LIDAR or Ultrasonic rangefinder which as accuracy of 0.5% at 5m

[@berniejansen5711]

Hi can you measure the depth off a drilled hole into the surface of the earth using the lidar sensor?

[@AndreasSpiess]

I can imagine. But maybe a range finder is sufficient?

[@prathameshdeshmukh7076]

I recently purchased the TF 02 pro from one of the resellers in the UK and the package didn't contain the appropriate connectors to hook it up with an Arduino, while your other model TF mini does come with that connector. Could you please comment on where could I get these connectors from?

[@AndreasSpiess]

I do not know. Maybe you ask your supplier?

[@annasarahkramerfazendeiro8030]

Would you recommend Lidar to take measurements between rows of trees (to maintain an accurate path between lines of fruit trees)?

[@AndreasSpiess]

If the distances are in the range and the trees not too thin I would give it a try.

[@annasarahkramerfazendeiro8030]

@@AndreasSpiess Thank you for your opinion! Might try it then.

[@kal9001]

Can a kinect unit be used for accurate range finding? I guess it would need lots of processing, and lots of power, but its view of view is huge compared with these measuring a single point at a time.

[@AndreasSpiess]

I think you are right. However, I have no experience with this device

[@3D_Printing]

some may want to measure small distances remotely with accuracy a bit like a micrometer at a distance no contact

[@AndreasSpiess]

I never tried that.

[@t1mmy13]

Maybe you can 3d print a base on which you can mount all the sensors side by side?

[@AndreasSpiess]

I probably could. But this was a one-off test. So it was easier to use an old piece of wood I used already for other distance tests...

[@t1mmy13]

Ah of course, K.I.S.S.! Keep It Simple Stupid! :)

[@EvenTheDogAgrees]

The phase detection could be made faster by using different colour lasers in parallel.

[@AndreasSpiess]

Good idea...

[@beforebefore]

Running multiple LIDAR devices simultaneously, right next to each other, at close range (1m) is cause for interference on sensitive units like the Garmin... it's watching for return pulses, and it may be getting false returns from other transmitters.

[@AndreasSpiess]

You are right.

[@tombouie]

I hope to use LIDARs for positioning on CNC. I'll need to put laser reflectors on the spindle. ?Are laser reflectors makeable/available for LIDARs? Thks

[@AndreasSpiess]

I cannot imagine the accuracy is good enough for a CNC. Lidars use normal lasers.

[@alexz9365]

Have you tried any sensors from Lightware, e.g. the SF000/B or the SF20/C?

[@AndreasSpiess]

No. It seems to be a new product.

[@kaikart123]

Do you know any LIDAR with reasonable price to be able to detect up to 1 km or there isn't?

[@AndreasSpiess]

No

[@SidneyCritic]

I wonder what's up with that Garmin, like is it super sensitive so it needs software averaging, or is it just not very good.

[@AndreasSpiess]

I think for obstacle avoidance the precision is ok. And its range is very long compared to the rest.

[@suicidalbanananana]

My guess is that it has the best IR detector, and thats why it sometimes still see's sunlight in the above test setup, causing it to get confused and sometimes just output a wrong range all together. Like Andreas said though, they are pretty far off numbers, so it shouldn't be to hard to filter those out/work with averages of several measurements. Edit: Another thought i just had, as its apparently auto-configuring itself at set intervals, maybe thats the cause, the way off returned results are from the moment when the sensor goes into 'setup mode' for a few milliseconds? that would be an easy fix (if interval is met then auto configure but dont return results, else return results)

[@SimenSupreme]

could the wierd value of the garmin be related to the auto calibration?

[@AndreasSpiess]

No, it was not connected to that. Calibration was done every 100 measurements and these values came randomly.

[@chuxxsss]

Look a lot like radar, same principles. Plus reminds me of radio theory. I don't know how you find the time Andreas to do all this stuff.

[@AndreasSpiess]

Me too...

[@moses1202]

An actual ToF sensor doesn’t measure the time difference. It measures the phase shift by IQ-modulation. I’ve been working on calibration of TI’s ToF sensor for around one year. It produces 10MHz as modulation frequency and mix it with reflected signal which suppose be similar in shape to the modulation signal but weaker. Through IQ modulation you can get very high resolution I phase and Q phase results, where the resolution depends on the strength of signal and resolution of adc. With I/Q phase, you can get the actual phase shift, and the phase results will be corrected by linearity correction cause the wave is squared but the signal through IQ phase conversion is supposed to be sin wave.

[@moses1202]

The calibration of ToF sensor is much more than calibrating distance offset. It have some characteristics change with temperature or ambient light. That’s a bit complicated.

[@AndreasSpiess]

Thank you for sharing your insight. I do not know which sensors still work with the indirect method (signal modulation) and I did not test if I see a modulation on the emitter.

[@matthews4159]

Xiaomi Vacuum Cleaner Robot Review and Teardown kzbin.info/www/bejne/rZjPe61oaKynatk 1of2 kzbin.info/www/bejne/aqS0dK1taqeZqK8 2of2 lidar/ir view from security camera kzbin.info/www/bejne/Z2SsoniOi9SNeqc

[@AltMarc]

The 1500m is more a matter of safety, how powerful is the laser in these units?

[@AndreasSpiess]

It says "Class 1"

[@uwezimmermann5427]

the Sharp analog distance sensors use triangulation and not the intensity.

[@AndreasSpiess]

You are right! I did not read the datasheet correctly!

[@uwezimmermann5427]

I'm using this sensor in a lab with my students...

[@benvanviegen2725]

Does anybody know where to buy such a phase detection sensor? Because I would like to use one with an Arduino, but I can't seem to find such a sensor.

[@kottelkannim4919]

Hello, Have you found one since? Thank you.

[@matze1508]

have you looked into the LIDARs of cheap chinese vacuum robots? might be a little bit more expensive but promising alternative

[@CutterSlade001]

I also wanted to point that out: www.aliexpress.com/item/Spare-part-Laser-Distance-Sensor-LDS-for-Xiaomi-Mi-Robot-Vacuum-Cleaner/32879027257.html?spm=2114.search0104.3.29.14ac4de7GzZrMO&ws_ab_test=searchweb0_0,searchweb201602_1_10320_10152_10151_10065_10321_10344_10068_10342_10547_10343_10322_10340_10548_10341_10193_10696_10194_10084_10083_10618_10304_10307_10820_10821_10302_10059_100031_10319_10103_10624_10623_10622_10621_10620,searchweb201603_2,ppcSwitch_7&algo_expid=a166696e-dfee-4394-8201-0a79d6a87c0d-4&algo_pvid=a166696e-dfee-4394-8201-0a79d6a87c0d&transAbTest=ae803_2&priceBeautifyAB=0

[@AndreasSpiess]

Maybe it is very similar to the one I have. Same look and same price.

[@MrZetor]

Good find! I didn't even know that the cheap Chinese robo-vacs had LIDARs already. I was actually going to suggest scavenging used and/or broken Neato or Vorwerk robo-vacs. They (at least the high end models) have featured a LIDAR at least for five years, and from what I have read, it is very good and well documented by the hacker community.

[@paulie-g]

It uses triangulation and there's a lot of community work available for it. I've got a vague recollection of seeing them on ebay for less from time to time.

[@ShivaniSingh-jc4pg]

i need your help to make my own laser lidar sensor of high range with low power consumption

[@AndreasSpiess]

I do not know how to do it.

[@coyote20030]

I would like to see the data from the Garmin. How bad was the deviation?

[@AndreasSpiess]

It was pretty much as described in their data sheet

[@coyote20030]

Gotcha. I'll take a look. Thank you..

[@connecticutaggie]

Did you find any discrete phase shift sensors? I only saw the packaged electronic ruler. This would not fit well in an embedded project very well. Regarding the sampling/measuring speed It would seem that this approach could be slow to find the target; but, once you found the target, you could measure quicker and infer the location possibilities based on its max/measured speed the time since the last measurement. Isaac Newton can be your friend.

[@AndreasSpiess]

You are right. But these algorithms are usually quite deep in the sensor chips. If you want to change that you have to build your own sensor from scratch.

[@connecticutaggie]

I am sure you are right, but not for cheap (unless you know a way). I need a sensor that can measure a soft target travelling about 5m/s, 0-40m range. I want to calculate speed vs distance, 10% accuracy and 300ms sampling may be enough; but, of course more is better.

[@ted_van_loon]

right now a laser depthmeasuring device with a range of around 50 meters and a measurepoint size of around 1mm(a actual laser), and a display costs around €15. and a sensor for DIY projects more for less range and a super broad measure point. I wonder if there really isn't a very cheap laser tof sensor which actually really measures at a laser fine spot, so with a super fine angle so that it can actually be used for scanning.

[@AndreasSpiess]

My problem back then was the lack of an interface for my IOT devces

[@berniejansen8701]

Hi, can the lidar sensor work in very dark environments?

[@juunas500]

Yes, the darker and the less ambient light there is the better a lidar works, since lidars rely on their own light source (laser) (, and only measures the light of its own laser).

[@lakshmipathibalaji2023]

@@juunas500 but dark surfaces will probably absorb the the lights...!

[@juunas500]

@@lakshmipathibalaji2023 Yes if the material of the surfaces is light absorbing, very dark black matte painted room for example. But if it's just an unlit concrete hall there should be no problem

[@indramal]

Do you know any software can be use to get 360 map for GP2Y0A21YK0F senor or TFmini sensor with arduino? I can use servo motor or stepper for rotating.

[@AndreasSpiess]

It is simply a analogRead command an the formula in my chart for the distance. Maybe you even find a library. Like that you can include it in any other Lidar projects

[@indramal]

I am looking for 2D or 3D visualizing software for GP2Y0A21YK0F senor or TFmini sensor with only programming arduino.

[@AndreasSpiess]

I never say something like that. Most people used Processing at the PC for visualization and an Arduino for capturing the data. Here is a project of one of my viewers: github.com/Brezensalzer/StepperBotLidar2

[@indramal]

Yeah you never say that but i am looking for that. that is why I asked. thank you very much.

[@AdityaSingh-in9lr]

Hey if you are looking for affordable LIDAR or a complete ROS compatible Hardware kit for Autonomous Mobile Robot you can check centaurirobotics.in/dev/

[@IDoNotLikeHandlesOnYT]

Here's my very short and to-the-point rant against the capitalization of "lidar" (as "LIDAR" or "LiDAR", both commonly seen): 1. You don't capitalize the analogous acronyms "radar" and "sonar" as "RADAR" and "SONAR" or "RaDAR" and "SoNAR", do you? 2. When the word was coined, the very first times it ever appeared in print, it was printed in all lowercase.

[@gilfoyle46]

Will it measure the distance of an object inside a tube?...Or will there be an error due to reflections from the walls of the tube?

[@AndreasSpiess]

There should be no reflection coming back to the device. The walls are in parallel to the beam

[@gilfoyle46]

Thanks. Actually that happens to be the case with ultrasonic sensors and IR sensors.

[@AndreasSpiess]

Lasers are very focused. Ultrasonic devices have an opening angle of 30 degrees.

[@dobrolv]

From your conclusions, TFMini can be used underwater? What about distances and accuracy underwater? As far as I understand, the frequency of the lidar is crucial here.

[@AndreasSpiess]

I have no idea if they would work and how far the LED would reach in water.

[@dobrolv]

I tried to put Garmin Lidar in the tank and get the measurements underwater, but it seems, that glass and water makes strong reflection surface. As far, as I looked through the internet, green lasers are being used underwater, because they have different wavelength. Lidar lite was NIR laser, however, underwater lidars are close to the green spectrum.

[@Jinguapingi]

13:06 - You lithium battery seems swollen ?

[@AndreasSpiess]

My batteries are usually swollen. You are right.

[@marslanthaheem422]

Hi i want to make a project written below I want to make a detector for detecting racing cars like after how much time a car has reached the ending point or when did it start running from the starting point. So I would want a sensor that can be placed at the side of the track (doing measurements on cars from the side view). There are two options that I found > Lidar > IR (placed on both sides of track, sender, and receiver) I would want a sensor that can >Detect 20-30 meters range >Sunlight generally does not affect it >Rain should not affect it >have long term use >Economical >Works with Arduino/Rasp pi ? > Should be able to detect cars and 4 wheelers I am okay with both Lidar and IR but when I was searching the internet I couldn't find IR with this much range can any one help me?

[@AndreasSpiess]

I am a KZbinr and provide my knowledge in my videos. The rest is on you :-(

[@marslanthaheem422]

@@AndreasSpiess thanks for your reply :)

[@Vidicon31]

The RPLidar uses triangulation.

[@AndreasSpiess]

Where do you got the info from?

[@Vidicon31]

Andreas Spiess The RPLidar is based on the Lider from the Neato robotic vacuum. If you disassemble the lidar you find a ccd array sensor. I have not done that because it destroys the calebration of my sensor. But there are fotos and documentation online. hackaday.com/2011/09/12/digging-deep-into-the-neatos-lidar-module/ As sensor it uses the DLIS-2K CCD dynamax-imaging.com/products/line-scan-product/dlis2k-2/

[@AndreasSpiess]

Thank you for the link. Unfortunately the video and the link to the blog are removed. Do you know why?

[@Vidicon31]

Andreas Spiess No I don't know that. But you can find patents on the system. patents.google.com/?assignee=Neato+Robotics%2c+Inc.

[@f-estoconingcasseyl.1993]

How much does lidar cost?

[@AndreasSpiess]

Usually, you find the links in the video description

[@danielegger6460]

I don't see any reason why phase based metering must be slow. In fact this has been done by SLRs for ages and is quite fast there (depending of your understanding of "slow", that is).

[@AndreasSpiess]

Fast here is a few 100 full measurements per second.

[@AndreasSpiess]

And I am also not sure about the max distance it needs. Maybe they can stop at a relatively short distance and therefore do not need many different wavelengths?

[@danielegger6460]

Seems more like a limitation of your product than the methodology, there shouldn't be any problem to reach 100 measurements per second with phase based metering.

[@danielegger6460]

They do have to work at the full range of minimum focusing distance to "infinity" and modern cameras can do dozens of frames per second with full subject tracking using many sensor fields, so clever processing is required within that timeframe as well. The only thing I'm not sure about is whether they can trust that they found the correct distance or they do a verification measurement that the lens focused at the right spot or even do it incrementally... Anyway. Sounds like an interesting research project.

[@AndreasSpiess]

I thought about our discussion and asked myself if they really measure distance or if they search for some patterns inside the picture. I assume their goal is not to know the distance to the mm but to create sharp pictures. So they might use a completely different principle than the one I described. But I have no idea about that matter...

[@prof.dr.rajeshghongade7224]

Can these lidars work underwater?

[@AndreasSpiess]

I do not think so.

[@AdityaSingh-in9lr]

Hey, Rajesh if your are interested to to make your own Autonomous mobile Robot you can check bit.ly/CK_9

[@Helli__]

I wonder if it would be worth it to salvage LIDAR sensors from defect vacuum robots. They are very cheap to get from ebay.

[@AndreasSpiess]

If you find the interface description they could be useful

[@Helli__]

@@AndreasSpiess I think I try to get one for a certain price limit I set my self. Just to look inside, and if the lidar sensor can not be used at least there are other parts to salvage, like motors, wheels, etc..

[@JimSims1952]

As always, nice video, thanks. I would like to measure the tide level in the salt water canal in my back yard. I had success with an ultrasonic sensor (HY-SRF05) but I am concerned that the salt atmosphere will destroy my project. I have tried to protect the project with a Saran Wrap window in an enclosure but the signal is attenuated and will not give a reading. I have recently tried a Sparkfun product number 14667 (sparkfun.com/products/14667) for $24.95. It uses the VL53L1X. I connected to an Arduino Fio which operates at the required 3.3 volts. They have a library and examples but because it is one of their SparkX "experimental" products there is no support. When connected to SDA (pin 4), SCL (pin 5), 3.3V power, and ground. Using the distance example, I only get zeros under test conditions in my garage. I asked about this on the Sparkfun Forum, as recommended, but I have not gotten any help. Andrew, maybe you could check this out and see if you can get it to work.

[@AndreasSpiess]

I made a video with this chip. Maybe you watch it

[@AndreasSpiess]

I just saw you have the newer version. So I have no experience.

[@JimSims1952]

I will hack around a little more and find out if I can make it work. I will switch to a "regular" atrduino" with a level shifter. I will keep you posted.

[@sourcesource8750]

Instead why not use mmWave sensor ?

[@AndreasSpiess]

what is mmWave?

[@sourcesource8750]

Andreas Spiess milli meter wave, frequency above micro wave, Ti and Nxp have IC chips as radar soc, which are not so costly, but u cannot get accurate measurement

[@AndreasSpiess]

Do you have the names of the chips?

[@sourcesource8750]

Andreas Spiess pls check this, simalar to this NXP and AD also has www.ti.com/sensors/mmwave/overview.html

[@ShivaniSingh-jc4pg]

how to make my own lidar laser sensor

[@AndreasSpiess]

I do not know

[@AdityaSingh-in9lr]

hey, you can now buy Lidar at affordable rates in India, check out centaurirobotics.in/dev/

[@nathiral-anasweh2609]

Dear sir I have been trying to use a Bosch 3 wire sensor that I got from used cars parts the sensor will be best for outdoor use The sensor is : Bosch 0 263 023 939 Parking Assist Sensor I hope you can help Natheer Anasweh, Jordan

[@AndreasSpiess]

I do not know this sensor. So you have to google to find information (if it exists).

[@3D_Printing]

Light interference patterns

[@AndreasSpiess]

??

[@GooGoosDad]

I'd like to have a Swiss accent and sound like a 24 year old!

[@AndreasSpiess]

:-)

[@hpmaxim]

Your explanation of the multi-frequency phase shift is not quite accurate. There is no reason to operate at a low frequency (say 10MHz) to get a coarse range, and a high frequency of 100MHz to get better resolution. Instead, select two high frequencies. Say 150MHz and 151.73MHz... Anything that is an integer multiple of a meter away would give 0 phase shift at 150MHz, so we know its an integer multiple of 1 meter, but which one? At 151.73MHz, you get a phase shift of 4.152 degrees per meter. At 150 meters, we get 0 phase shift on the 150MHz, and 262.8 degrees on 151.73MHz. By comparison, 63 meters is 261.576 degrees, and 64 meters is 265.73 degrees. 236 meters would be the next point where it comes close but its further than 63. Then 323 meters, but again, its further than 63 meters was. For additional clarity, third frequency could be added. This technique does require some amount of math since there is no closed form solution to figuring out the distance based on the phase shift of multiple frequencies, but computing power is cheap.

[@AndreasSpiess]

s this a theoretical idea or do you know that my device worked like that? I did not do exact measurements.

[@hpmaxim]

@@AndreasSpiess I certainly can't tell you I know as a fact how every phase shift range finder works. I found a Russian guy who had done a write up and claimed this was how it was done (I made up the exact frequencies, but I think he was putting them in the 150-200MHz range). As for whether it can be done theoretically versus practically. I'm pretty sure its quite practical. It'd be like saying you could use an XOR phase detector to do a phase error -> voltage converter, or use an FFT and look at the phase difference by taking the atan (imaginary/real) at the desired frequency. In fact, the Russian recommended using the Goertzel algorithm. It would be possible to modulate with MULTIPLE frequencies simultaneously since you're unlikely to see any optical mixing if you are using FFT or Goertzel which would make it faster. Of course, it also means you have a reduced SNR so you'd have to put out more power to compensate. What I can tell you is, that the method I described seems to be a significantly better option than using a high and low frequency, since a 10MHz wave still gives you a maximum unambiguous range of 15 meters, and these range finders can work out to hundreds of meters.

[@electron-1979]

Engineers don't use cm! Let's stick with Engineering or Scientific Notation. LeddarTech sell their LeddarOne for about $100 and is worth a look at.

[@AndreasSpiess]

I looked at the leddarOne. Looks good. I will put it on my list...