Ideas/Suggestions Needed

[Lunat1c]

Hi,

As an EE student I'm currently taking a class in which we're grouped up with a number of other people and we have to come up with a project of our liking to work on. Basically, my group is going to be making a camera dolly with vertical stabilization (vertical stabilization meaning, the camera which will be mounted on a vertical axis will move up/down automatically to always keep a pre-set height from the ground when one moves the platform on which everything is built up/down). The problem I have with this, is that I'm finding it hard to find the use of such a thing in real-life and I don't want to work on something which is completely useless. So we tried to make some alterations to the above mentioned so that apart from it being more applicable it will also be more challenging. 

The first thing we decided to change is that instead of having the camera move up/down only, the axes on which the camera is mounted will be able to rotate 360.

I also came up with the idea of having the camera be able to track something. That way, it will keep following something within a certain radius of it (the dolly remains stationary but the camera can rotate and move up/down as already described).

The reason I'm telling you all this, is so that maybe some of you can help me come up with different ideas on how I can make this more interesting. If any of the above mentioned ideas sound dumb or unreasonable, please feel free to mention so. I won't be offended. Basically, we have about 2.5-3 months to finish the whole thing so whatever is proposed has to be doable in this timeframe. Thanks a lot for any help you could offer!

[sacherjj]

Tracking a certain thing around ~360 degrees plus a certain amount of up/down would be interesting.  However, make sure you have the skill set for it.  This will require more software and image processing.  Your first idea seemed like it would only be gyro/accelerometer based or similar.

You could simplify this with a known target that was fairly easy to detect in software.  Then your electronics could interface with a computer using possibly just a web cam.  Then software becomes easier.

Alternatively, you could use a Kinect as your sensing platform.  That provides video combined with distance information in the IR grid that the Kinect uses.  There is also quite a bit of open source projects with good info on this interface.

[olsenn]

I strongly support the KISS (Keep It Simple Stupid) philosophy... no need to make things more complicated for yourself; the purpose of this design is to learn how to use motors and sensors... your vertical stabalization system will be plenty difficult enough to perfect.

As for how it could be used in practical terms... perhaps you may want to make a video someday without the camera being shaky (signs of a cheaply made film). I can think of many applications for this type of design; however, I think the biggest bennefit to this design is that you will learn how to use and tweak into perfection distance sensors, and control a motor in real time, minimizing latency and compensating for whatever latency does exist.

If you really want a challenge, make this circuit without using a microcontroller.

[caroper]

The problem I have with this, is that I'm finding it hard to find the use of such a thing in real-life and I don't want to work on something which is completely useless.

You must be a landlubber

Such a device would have lots of uses aboard a floating platform of any kind.
Possibly the most important would be navigation and auto pilot functions, it should also prove useful in sea rescue situations. Another use would be documentary or research photography of such things as polar ice flows or marine animal populations.
Don't dismiss it as useless just because you can't conceive a use in your normal frame of reference, think outside the box.

I agree with tho other posters, KISS, 3.5 months is not a lot of time especially for a system that crosses several engineering fields, Mechanical, electrical, optical and software.

Rather aim to do an excellent job of a basic problem than a reasonable job of a complex one.
You can always expand the scope once you have submitted the results and have learnt the limitations.

[Lunat1c]

Thanks for the input. I really appreciate it. For the above mentioned reasons (and a couple other reasons as well), we're gonna stick to this idea. Now one question I have is, what sort of motor is best suited for this application given that a very fast system is required?

[HLA-27b]

Voice coil actuators like this one are pretty fast.

[bfritz]

When choosing the electromechanical components, you need to have some idea of what the mechanical requirements are.

So lets think a little about the actuator that would move the camera up/down.  How much does the camera weigh?  How much will the rest of the assembly that you move up/down weigh?  How much do you need to be able to move the camera up/down?  This things will have an interrelated effect on the actuator you choose.  For example, let us say you have a 1lb. camera, and need to move it +/- 1.5 ft.  The easiest solution is then an actuator rated for a the weight of the camera, plus the mounting assembly.  Note that you will likely need to increase the force capability as you need to think about how fast you need to move the camera up/down.  Remember that F=ma, so if you need to accelerate / decelarate quickly, the force requirments of the actuator go up.  Then once you know the force needed, then you can think about the travel.  In this case you need a total travel of 3ft.  My guess is that an actuator capable of the force required, at a 3ft travel, may be sort of expensive.  You may then say, well lets use mechanical advantage with a linkage that then increases your travel.  Note that this will likely force the total weight of the camera assembly up, and remember that if you then design a mechanical advange of 3, that the forces from the weight of the camera and platform on the actuator will be increased by at least 3x, and likely a bit more, as no system is 100% efficient.  (Our old friend friction comes into play.)

So, you can iterate on a couple solutions, while looking at total cost, in an effort to minimize the total solution cost, so that you can build the thing.  Then you can come up with a best guess on where the sweet spot is for cost of that solution.

Do this in other areas as well.  Do you need to know the position of the actuator?  How will you do so?  How will you determine if the camera must go up/down?  Can you determine if your feedback system will implement a proportional response, or will it be a up-off-down control loop?  etc.

My guess is that you will need to find specs for a variety of types of actuators, with different force, response, and travel requirments, and then use that to choose the device.  Note that the optimal solution may even include one device with a long travel with slower response, and a shorter throw device with fast response.

There are many more variables here than you've thought about.  This will be plenty challenging for you!  In engineering, lots of times you aren't the expert... until the end of the project where you have become the expert!

[Rerouter]

Or if you want to get away with a weaker actuator, you can mount the camera on its balance point, then the only force needed to be applied is to overcome its inertia,

as an upside if balanced well, when no power is applied the system will remain where it was placed.

as for tracking, if you want to keep it nice and simple, make use of an IR emmiter and an array of 5 phototransistors, 1 center, and then 4 around it, with some flaps to limit there feild of veiw to about 15 degrees, or less, and that way you can do tracking with a micro, and it will know when its centered as only the center one should be on due to the flaps blocking the others,

[DaveHardy]

These guys have pretty much set the bar for this type of thing.   It's self balancing technology based on gravity and inertia.  I doubt that you could beat it with a microprocessor.   Operating one of the things is an art in itself and the camera operator guys make huge money.

http://www.steadicam.com/

Check out this awesome video:



You know you're good when you've got people filming you filming other people!!! - The top YouTube Comment!!  Ha Ha

[bfritz]

Or if you want to get away with a weaker actuator, you can mount the camera on its balance point, then the only force needed to be applied is to overcome its inertia,

Note that this will not accomlish the goal stated.  What you are proposing is an actuator that tilts the camera up/down.  This is much different from the goal the OP has given...

Basically, my group is going to be making a camera dolly with vertical stabilization (vertical stabilization meaning, the camera which will be mounted on a vertical axis will move up/down automatically to always keep a pre-set height from the ground when one moves the platform on which everything is built up/down).

Here the idea is to allow the camera guy to set the tilt, which is a seperate axis, from being able to move the camera up/down to keep a fixed and constant altitude.