Carbon Fiber for outside project

[Woody1458]

I know it is really close to the season to start an outside project, but opportunity causes the inspiration not boredom. I am building a robotic arm for a guy that will allow him to autonomously take panorama photographs perfectly everytime. The device needs to obviously then be able to hold its own under normal circumstances and not flex more then a millimeter. Also we would like the device to be backpackable so that it can be used in remote locations for great shots. (sub 30 lbs) Also it needs to be corrosion proof, so It can be left in position for days maybe months at a time. (taking pictures of Eagle babies). Also small so that again it can be fit in a backpack. Also semi cheap, but that's not that important, because it is a high-quality device. Also durable so that it can be dropped while traveling yet continue to work for a long time.

So far I think Carbon-fiber is the best choice. I have no experience of carbon fiber other then a few hours of research I have done online. If I understand the process right I believe my best choice for production would be to craft the piece, in this case an arm tube, in a light yet stiff foam then wrapping it in 5 layers of carbon fiber then a process called "Vaccum infusing" applying the resin. I think I would then leave the foam in for solidity and to save time on the struggle of removing.

Please tell me if there is a better material to use, or a better way to manufacture carbon-fiber. Thanks in advance CD.

-Woody

PS the website of the guy who Im building this for is http://russellschutte.com/

You're right with the choice of CF. I'd love to give you some tips about working with it, but I learned the bits I know while under an NDA that still applies. However, you will find some help here. http://www.roarockit.com/ I suggest emailing the guy named Ted there, he's pretty big in composites use for longboarding. He'll definitely have some good advice for you. And they sell vacuum presses that work quite well.

[dlavery]

With all due respect to the desire to build something cool (I know that I succumb to it occasionally), this is one of those situations where you have to consider the best interests of the customer. In this case, there is a solution already commercially available that is designed to perform this exact task, and can be obtained for a price that is likely less than the cost of your raw materials. Have them take a look at the Gigapan Imager. This device is used as the standard device by the Gigapan Project for acquiring multi-billion pixel panorama images.



-dave





.

[Woody1458]

Craig, thank you for the reaffirmation. I will check out that website and toss an email twards Ted. Also do you open with women talking about that? Because the whole NDA thing seems hard-core

Dave, thank you for the reminder. When Russell approached me with his idea he included a 90 page research report on various things such as: engineering ideas, purposes, requirements, explanations, and most importantly to your comment: market research. There were, if I recall 4 similar devices he was able to find, including the Gigapan.

Pretty much in his segment on the Gigapan is that it is a consumer level device, not a professional level and that to his full understanding the project was scrapped before reaching its full potential. My next post will include his view on the Gigapan.

[Woody1458]

As promised, Russell's segment on the Gigapan:

GigaPan:
The GigaPan originated, apparently as a collaborative effort “sponsored by Google,
Carnegie Mellon University and the NASA Ames Intelligent Robotics Group, the Global
Connection Project.”

Despite this wealth of knowledge, the original GigaPan suffers from an inherent design
flaw – it rotates the camera at the film plane, not at the optical center of the lens, so it will
fail to create panoramic/gigapixel images with near foreground and distant background
subjects with SLR cameras (anything with a lens that protrudes far from the rotation
point). It may be possible to correct some parallax issues through very advanced image
seaming, but true photographers know you shoot your best images in the field – don’t
rely on software of any kind to fix issues you should have compensated for in the field.
More information about this issue, known as parallax, is explained in the Feature
Justification for Panoramic Photography.


It seems that they are now attempting to go commercial with this product, aimed at
consumer level buyers, in partnership with Charmed Labs
(http://www.charmedlabs.com/index.php...sk=view&id=111) but
their “Beta Testing” program closed October 2007.


I was selected for the Beta program but never went through with the evaluation since I
knew the product would not live up to the standards required for my art. If ever fully
developed, the product is rumored to sell for about $200 – I can still purchase as part of
the beta program for $349.

While they show the product with a Canon camera which looks similar to SLR cameras,
last time I spoke with them, they indicated that SLR cameras are not supported –
rendering this a fun hobbyist tool, not a pro photographer workhorse.

The GigaPan group has created software for viewing and sharing such detailed images at
www.gigapan.org. Charmed Labs also used to make a motor control brain product called
the Qwerk, but it’s no longer available on their website – it appears to have been
purchased by Innovation First, the makers of VEX Robotics.

The important part of this piece is in the second paragraph. Due to the device positioning every camera in the same spot regardless of the sensor and lens positioning it does not allow a photographer to take pictures of things in both close and distant range without the magic of the stitching revealing itself.

[Woody1458]

The product he found that is on the level we are looking at is

http://www.peaceriverstudios.com/pixorb/index.html

which costs $11,500.

The point of this project is that we think we can make a better machine that costs less than this.

PS: sorry about doing three consecutive posts, it just felt more smooth to do it this way.

[dlavery]

Unfortunately, the critique of the Gigapan Imager, while nicely detailed, is inherently incorrect. Perhaps if Russell had actually participated in the beta test program rather than giving up on it before he even put his hands on a unit, he might have been able to provide a more fact-based analysis.

For example, the criticism of the placement of the camera detailed in the second paragraph is substantially flawed. The positioning of the camera on the unit is adjustable, and the optical center of the lens can indeed be placed at the center of rotation. In extreme cases where long lenses or macro lenses are used, alternate baseplates are available that can be used to position the camera properly. Ron Schott's work photographing crystalline geologic structures provides some great examples of this capability.

Likewise, the statements in the fifth paragraph regarding SLR support are incorrect. There are versions of the device that are specifically designed for professional-level SLR support (I am getting one for my Nikon F-4, N-90 and D-200 cameras). Additionally, environment enclosures are available for ruggedized versions of the system that allow the device to be left unattended outdoors for months at a time (e.g. PBS has been prepping to use one of the units in the Amazon basin to acquire panoramas of the rain forests; this system is going to be left alone for up to a year in one of the most mechanism-unfriendly environments anywhere).

As to the comment in the preceding post stating "the project was scrapped before reaching its full potential," I do not know where this information came from. The Gigapan project is still very much active and ongoing, with funded development of additional versions of the tripod head underway.

There are a number of fundamental errors in the critique of the Gigapan. Russell may have been willing to accept this flawed analysis without serious question, and may have his mind made up regarding the solution he seeks. That is his prerogative. However, I do not believe it would be proper to have incorrect statements like these be displayed publicly here without being addressed.

-dave



.

[dlavery]

Originally Posted by Woody1458

The product he found that is on the level we are looking at is

http://www.peaceriverstudios.com/pixorb/index.html

which costs $11,500.

The point of this project is that we think we can make a better machine that costs less than this.

I have one of the manual Peace River Studios rotating tripod heads (the 3Sixty model). The construction is very high quality, and the design of the unit is fairly well thought out. The only critical comment of any type that I have for it is that the rotation step increment is built into the hardware, and cannot be altered (which is necessary for different lens lengths). This is something that they seem to have addressed with the PixOrb HD model. Other than that, the 3Sixty is a great unit and has served me well.

If you are looking to replicate something with a similar level of construction as the Peace River Studios solution (such as the PixOrb HD), you should be prepared to spend a significant amount on the development. Creating a one-off unit with a similar level of robustness and quality for a lower price point will be a significant accomplishment.

-dave



.