High resolution short range distance measurement of a moving object

Could you recommend a distance measurement sensor? The sensor will be used to measure spread of a concrete. See an example here:

https://www.youtube.com/watch?v=x3JATc7gho8

The idea is that I will place the sensor on the table next to the cone and point the sensor towards the concrete in the cone. The cone gets lifted and the concrete flows. The sensor will point to the concrete from side and measure the distance of the concrete from the sensor while the concrete flows. The result should be distance of the concrete front to the sensor as a function of time.

The requirements are:

  • The sensor should work with fresh concrete

  • The sensor should measure the distance at the height approx. 0.5 cm above the table. The concrete layer gets very thin at the end so it should be possible to hit the concrete from side even when the layer is very thin => therefore approx. 0.5 cm

  • The sensor should work in the range 20 - 50 cm

  • The precision should be roughly 1 mm

  • It should measure the distance min approx 5-10 times per second

  • The price should be roughly below 1000 USD

  • The unit should have USB or similar to connect it to laptop

  • It should be possible to buy it in Europe (Denmark, Germany...)

  • It should work outdoor

Any idea what to use? Should I go with Ultrasound, infrared, laser or anything else? Can you recommend any particular device?

If you are looking for an off the shelf solution I wouldnt know of any...sorry but if you are looking for a project...then...

If I understand correctly, you really need to know the distance (in real-time) from the outer circle of the poured concrete to the edge of the platform which is within 50cm of the centre, with a view to calcuting the 'flow' rate for some sort of QA.

My guess is that it would be difficult to achieve for reasonable cost in the setup you have shown.

Typical sensors would be Ultrasound or IR. I'm not sure US would be sensitive or accurate enough. IR would suffer from reflections and possible ambient interference.

Maybe a (hacked) laser measuring tape tool could do the job with minimum fuss? (Assuming the edge of the wet cement provides enough reflection!)

Other approaches could include: - use a camera overhead and analyze the image in realtime against pre-calibrated measurments. (possible to do with RPi setup or laptop & basic usb camera) - insert moisure sensors at set intervals on one or more radius/diameter and extrapolate the speed of flow between sensors. - Embed some sort of 'resistive' wire(s) across one diameter and measure changes in resistance/voltage/current during flow?

Finally, here is a cool hack which shows what is possible (intercepting the screen display & sending out over usb!). Also look for meters with bluetooth or usb interfaces already, which would avoid the hacking. http://porcupineelectronics.com/Laserbotics.html

Sorry for the delayed response.

Yes you are correct. I need "position" of the concrete front (outer circle) in real time.

I was looking at various options. Something like Sick DT20 Hi would not do the job? E.g.

https://www.mysick.com/partnerPortal/ProductCatalog/DataSheet.aspx?ProductID=48107

What are the drawbacks of such a solution? The only problem that I see is that I would have to tilt the device to side so that it can shoot the laser at the height of 1 cm above the ground.

Currently, we use camera but all in all it is quite a complicated setup. Since we would like to simplify the setup we are looking for an alternative to the camera solution.

The ideas about the moisture and resistive wire are interesting. I will think about them ;)

I dont think that your approach would work very well (based on my understanding)

You could use this type of sensor mounted on a drive (like an x-axis of a CNC or 3D printer) and it is moved by a stepper motor in line & along with the flow.

It would be possible to make cheaper sensors, but if you need reliability in an industrial environment then probably worth paying for a robust sensor & mounting.

Overall a camera would be my preferred approach.....get/sponsor a good student to do it for you on a RPi if you think its too complex or just put the effort into simplifying the issues with the current 'complicated' setup.

Well the idea is that the device will be used in practise by concrete producers to measure rheology of the concrete. Currently, they have their own cones they use to lift on their own tables. They do not want to use any scientific setup :).

If we wanted to have camera above the cone then you need to have an arm to hold it etc. and it ends up being a huge device. Nobody will use that. Therefore I prefer a solution of a small measurement device connected to tablet that you place nearby the cone. Then you lift the cone and the tablet will provide you with the rheology of the concrete.

So what you say is that if I place the SICK device at a distance of 500 mm from the cone and point it towards the cone, I will not get good realtime distance measurements? Or is there complicated setup or calibration of the SICK device before every use? Or where is the catch?

I dont really know a lot about that sensor, but I dont think it will work well for your application.

Now that you say you have a tablet available - another idea springs to mind.

Place a black/white strip along a radius or diameter (like a long bar code). Mount the tablet on a tripod (readily available & needs to be steady).

Write an android app to measure the flow rate - as you can calibrate against the known dimension of the black/white strip. (This would be very similar to bar code scanning but at larger scale)

In fact just capturing images, with timestamps, as fast as possible and processing later offline would also be possible. Calculations could even be done manually.

Well, this is actually a very clever idea! I will think about it deeply :) Thanks a lot for your help!

ir DISTANCE MEASURING SENSOR

arielmanaligod: ir DISTANCE MEASURING SENSOR

Are not accurate to 1mm

Grumpy_Mike:

arielmanaligod: ir DISTANCE MEASURING SENSOR

Are not accurate to 1mm

What is the usual accuracy?