1
$\begingroup$

I'm currently designing an autonomous robotic system to manipulate clothes using computer vision and complex moving hardware. My current design incorporates quite a number of moving parts. My biggest worry is a frame (140 x 80 x 40 cm) rotates from 0 to 90 degrees every time it manipulates a piece of cloth. Other than this the design involves various other moving parts to achieve successful manipulation of the cloth. It seems like the hardware is capable of achieving the task despite the high number of complex and moving parts.

So the question is, what are the design considerations I should take in designing an automated system. Should I think of a alternative design with less number of parts? Or proceed with the current design if it does the job>? Sorry I am in a position where I can't disclose much information about the project.

Thank you.

$\endgroup$

3 Answers 3

3
$\begingroup$

Tough to say without knowing much about the project. More parts = more problems or potential points of failure. A simple design is typically better in that there are less things that can go wrong. If the precision you need can't be accomplished without the extra parts you may want to reconsider the design.

Can you accomplish the same thing a simpler way? Will the parts need to be calibrated? Will downtime for repairs be a significant obstacle to the adoption of your device? If you are planning to bring it to mass market, how difficult will it be for someone to repair it? Can the machine self correct in the case of calibration. For true autonomy the machine should be able to make these adjustments as needed and if you are using quality parts and the system is engineered well these things become less of an obstacle.

Only you can answer these things.

$\endgroup$
0
2
$\begingroup$

“Art is the elimination of the unnecessary.” - Pablo Picasso

Many designs can meet the requirements you set. To prove it works is one phase. To optimize is another. What optimize means depends on your priorities

When you understand your requirements then specs, you should consider many design philosophy points like simplicity, modularity, flexibility.

I would recommend The Art of Unix Philosophy as a set of koans, about what priorities you should consider. Even though it's software, it's same system modularity. Also, it's a free pdf.

Mechanical systems now fail on lifetimes versus electronics every couple of years.

Second, you can also consider how many COTS off the shelf parts versus designing your own. Why redesign a part if you can use one already? Can you reduce the number of parts by changing one into a common one with a second set of holes? What about a common module? Lots of ways to make design better.

But before you design / redesign anything, ask yourself what is the value of the redesign? Should you do that now or wait till you understand your system better?

Any more advice let me know.

$\endgroup$
1
  • 1
    $\begingroup$ Interesting quote, reminds me of "It seems that perfection is attained not when there is nothing more to add, but when there is nothing more to remove." - Antoine de Saint Exupéry $\endgroup$ Aug 24, 2016 at 7:52
2
$\begingroup$

Simplicity is the greatest sophistication.

Keep the design simple to understand, implement, debug, and maintain. Do not have too many moving parts. First, start with a simple design and build more functionality on top of it.

$\endgroup$
0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.