Measuring speed, angle and height

Hi Salvador,

to be honest, I think that with some difficulties due the limitations of the Arduino board it is not a so complex operation. There is some math and the sensors should be chosen accurately in my opinion, just to be sure that yuou can retrieve the data quality and precision that you need and expect.

Do you have some idea of just only the math vision of the process you should calculate? Do you need some suggestions on the sensors ? Maybe also interesting to know what is your expected budget.

Also important some more details on the object type, fly speed, typical path (flying or going down or what else?) Etc.

Enrico

Thank you Enrico. Personally I thought that this would be almost impossible. I am glad you say that is not complex. I am working on getting more information. At this time I am gathering more information, basically some of the same questions that you have. I do not know which sensor could reliably get the information required, Any recommendations are welcomed, but let me get the budget first.

And no idea about the math. One step at a time :-)

Best regards, Salvador

Hi Salvador,

impossible I think not, but with Arduino very difficult. The point it not Arduino as the "concept" but it most depends on the speed of the processor and its limits. Then about the math I have experienced in past some strong trigonometry to make some ground motion information to calculate speeds, acceleration etc and the limits of this board and its 16MHz are emerging.

you write that math ... after, one step at a time. In my opinion, this is the key actor and we should know first of all the kind of calculations, how much "real time" we expect to have results, what kind of data should be produced. Another point to take in account is "what is this flying object?" Another influencing factor is the powering of the system (battery weight), its size and other physical details that may or may not influence the creation of the project.

Another question;

measure the speed, angle and height of a flying object over a 6 x 30 cm surface.

The collected data are for logging purposes or should be used as feedback of the flying object? Do we expect that data are sent somewhere else wireless while the object is flying, and if yes at what distance? Or data should be available as processed for on-object immediate processing?

With the reduced information I have and the scenario that you depicted at the moment the only alternative and probably reliable platform I see to solve your problem is a PsOC4 or PsOC5 device that I am almost used to, with a great programming platform not difficult to manage. This low power device has also the advantage to be very cheap and powerful. I am not an expert in PsOC but I am sure this is one of the possible best solutions.

Let me know. Enrico

Thanks again, I appreciate your assistance. I have been asking questions:

The object: A free flying tennis ball travelling at approximately 70 to 90 MPH (110 - 145KPH) always on a positive slope

The budget: $50 USD

The math is interesting. I am looking for three measurements. The height can be taken care of with a distance sensor on the surface, measuring the distance as the object flies over head. The angle can be calculated by taking various height readings at different points of the surface and then calculating the slope and getting the angle. I'll admit my trigonometry is rust, but nothing some research can't fix :-). The speed can be calculated by the time it takes the ball to go from one height sensor to another.

Still, given the answers to the questions I've asked so far, to me it seems that it might not be doable. 70 - 90 MPH is fast and as you mentioned the 16MHz is a limitation. At the ball's speed, I am not sure if the device can take even one valid reading!

At this time I have only been concerned about the data acquisition, but your other design questions are very valid. What needs to be done with the data once it is acquired? I am going to ask that too, but have not because I have doubted the project's feasibility from the beginning and given the above parameters I am not sure it can be done with the intended hardware or budget.

Regards, Salvador

Must it be a normal tennis ball or can you modify it ?

How quickly must you come up with the result  after the ball passes ?

A vision based system is the obvious choice.

At 145kph = 40.3m/s the ball will take 1.489 ms to travel 6cm so you will need  a camera operating at 2000 frames per second or faster - lots of money.

Possibly, if the ball is always traveling fast, you might be able to use microphones to listen to it (this approach is used for tracking bullets) - you can't detect  slow tennis ball be sound because it is too quiet.

Leading on from this - perhaps you could "illuminate" the ball with modulated ultrasound and use several sensors to estimate its position.

The passive sound based idea could be cheap (for the sensors) but would need way more than Arduino processing power - but might get within $50 if you made 100k of them.

Active ultrasound would cost more  - unless you can train bats

MK

TI make a pretty good ultrasound chip you might consider.

http://www.ti.com/lit/ds/symlink/pga450-q1.pdf

and this might help too.

http://e2e.ti.com/blogs_/b/analogwire/archive/2015/07/31/time-of-flight-measurements-using-high-performance-time-to-digital-converters

Already though to reverse the problem? Tennis ball can be a good marker to be followed from a ground fixed (or mobile) point of view. If possible (depends on where and how the tennis ball is launched / landed) this can simplify enormously the problem.

A second intermediate alternative is the bare sensors + medium range radio transmission on the flying object and a ground unit collecting and processing data. The point in this case (but also with all the system onboard) is that the sensors changes the aerodynamics of the flying object and consequently its behaviour.

Enrico

Thanks MK!

I would assume a normal tennis ball.

Given the parameters of this project, training bats is looking more feasible. At this time I am still gathering information about the project. Will post as I get more info.

Thanks Peter, I will have a look at the links that you posted. Salvador

Salvador can't use a classical ultrasonic distance measuring system because the tennis ball moves much too fast - he has a total window of about 1.5ms as it goes past and he needs to measure it's position at least 3 times in that window. So at 0.5mS per measurement there isn't time to send a pulse and wait for an echo. (If the ball is 30cm away it's  a 2ms round trip time.)

Hence my suggestion of ultrasound illumination with a modulated carrier - but it won't be easy to demodulate the echoes because the Doppler shift will affect the return signal making it harder to lock onto the modulation (which you have to do to estimate the time from the transmitter to the tennis ball.)

(It's a fun project but I wouldn't look at it commercially for less than £50k for a working prototype !)

MK