color codes for identifiers

How about using different shapes like star, circle, square?

I think the surfaces will make that hard to perceive. (ignoring how difficult they would be to *create*).

And, adds to the *number* of things that someone would have to remember: black square, red circle.

It would likely be easier just to add a third "band" (which reintroduces the prospect of getting the bands out of order).

I thought of trying to use concentric shapes (like a red dot that fills a green circle) -- as it imposes some order and lets me increase the number of *permutations*. But, I can already imagine: "Matt, was that a red circle around a *green* dot or a green circle around a *red* dot?

It seems that just remembering two colors -- regardless of any "ordering" -- will be the least taxing.

So why not just use the resistor color code without any placement significance?

Or you could do something which makes the placement more significant such as body, tip, dot.

How many digits do you need?

Use letters and numbers.

Inkjet print.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

• Black-Brown
• Black-Red
• Black-Orange
• Black-Yellow
• Black-Green
• Black-Blue
• Black-Violet
• Black-Grey
• Black-White
• Brown-Red
• Brown-Orange
• Brown-Yellow
• Brown-Green
• Brown-Blue
• Brown-Violet
• Brown-Grey
• Brown-White
• Red-Orange
• Red-Yellow
• Red-Green
• Red-Blue
• Red-Violet
• Red-Grey
• Red-White
• Orange-Yellow
• Orange-Green
• Orange-Blue
• Orange-Violet
• Orange-Grey
• Orange-White
• Yellow-Green
• Yellow-Blue
• Yellow-Violet
• Yellow-Grey
• Yellow-White
• Green-Blue
• Green-Violet
• Green-Grey
• Green-White
• Blue-Violet
• Blue-Grey
• Blue-White
• Violet-Grey
• Violet-White
• Grey-White
• Black-Black
• Brown-Brown
• Red-Red
• Orange-Orange
• Yellow-Yellow

I've shown I can do it with 10 colors. But, that starts to put stress on the discernment of those colors (e.g., orange vs. red). So, I wanted to cut down the number of colors used AND the number of "bands" needed (again, easier to remember a combination of TWO colors that you've observed vs. FOUR)

Items marked are not consistent. So, you'd have to develop (and remember) a "system" that ensured you recounted the colors in the "correct order" for that type of item.

There is no significance to the colors chosen; no "digits" to which a mark/band/spot is mapped.

I just need to ensure a set of color markings uniquely marks a particular item -- regardless of their ordering (i.e., combinations, not permutations -- unless something else can ensure their ordering is conveyed in their recounting -- like yellow dot in red circle -- without adding significantly to the cognitive load... or increasing the chance of it being "misremembered" from one moment to the next)

Actually, it's answer made me realize that two marks of the *same* color could be a valid identification. Even if they "appeared" to be just one big mark (BLACK BLACK == BLACK ... if a single BLACK is not competing for that identity!)

Yes, but *10* colors. You can note the difference between red-orange (or orange-red). But, given orange-orange, can you (can THEY) determine that this is NOT red-red?

I've watched them visit the *sorted* bins of power cords and take one to see if it fits. Failing, return and select another. Etc. The idea of EXAMINING the mating connector not occurring to them in their selection process.

(why not take one of each, test to see which fits and then return the others??)

These intermittent failures add to their frustration as well as the time required to perform a task. Currently, requiring the "attention" of a supervisory person who likely could do the work THEY are doing in far less time!]

That was the reasoning behind my initial choice of red yellow black white green blue augmented by gold silver

I can get 72 "combinations" with these *8* colors -- but, at the expense of a third stripe (that, effectively, gives an implicit order to the *permutation* presented).

So, I'm trying to reduce the complexity of the encoding so they don't have to remember specific colors or specific orderings

In one facility, a gentleman would head off, down the ~100 ft hallway mumbling to himself (repeating the item that he was seeking from the far end of the hallway). Then, return a few moments later: "What was I supposed to get?"

...

If this is a case of mentally challenged people being given the opportunity to do these tasks then it sounds to me like the best solution would be for the supervisor to hand them the correct screwdriver or cable. Sure that takes longer but if they can't tell orange from red then it's hard to see a solution which does not have a supervisor assisting them.

You want them to learn to "take direction" (my understanding is that is the biggest part of the "exercise").

So, "put the scrap metal in the red-black bin" or "get a green-yellow cable" or "use the blue-red screwdriver".

Otherwise, they are just "passing time" and not really "progressing". The goal is to increase their ability to act on their own WITHOUT having someone there to tell them every detail about their life.

Consider how employees at McDonald's are "trained" to just push buttons labeled with pictures...

Ok that's fine as long as they are capable of learning.

Sounds like me using LTSpice.

I don't think the goal is to "teach them skills so they can hold down a job". But, rather, teach them how to do the things they will need to do to care for themselves (so a "supervisor" doesn't have to hand them a toothbrush and tell them to brush their teeth after a meal)

Some "higher functioning" individuals are "learning to be responsible" and show up at their appointed time, etc. They can likely learn a skill (even if it is a menial one) but need to learn to discipline themselves to "show up" (at work!) reliably.

Six colours, present/not present gives 64 combinations. Seven or more colours could be used so that there were always at least n colours present which may be preferable, or even to add a parity bit. Calculating n is an exercise left to the student.

Bit terse. Binary coded colours of course, position/size/shape doesn't matter, just presence or not.

eg Black = 1, Red = 2, Yellow = 4, Green = 8, Violet = 16, White = 32

We are exploring the mathematics of how many combinations of N things take M at a time there are.

.

Joe

Yes, that solves the "position/orientation independence" issue. But, it does so at a high cost (IMO).

- A user never knows how many colors he will be tasked with remembering (1 to 6)

- Potentially having to remember MANY colors

- Forgetting one of the present colors dramatically changes the "code" (with no way of the user realizing this)

- Remembering a color incorrectly (always a possibility, not just here)

- Conflicts between a color and the color of the surface on which it is painted (more later)

Who's going to "check" the parity/syndrome?

Here's the use case I was working on today:

I wired an "octopus" of barrel connectors to a single 12VDC brick/wall-wart (IIRC, there were about 30 different connectors -- though not all are technically *barrel* connectors) The brick is rated at 5A DC (larger than most anticipated loads). The brick has "12" painted conspicuously on its black surface. All barrel connectors are wired "center positive" (seems to be the norm in the US; I've seen center negative on some Sony kit and Siemens -- we discard those just to avoid the wrong polarity issue. Likewise for AC output supplies)

A user tasked with "checking" to see if an item (almost invariably encountered WITHOUT it's appropriate power supply/brick) is functional tries to mate each of the barrel connectors to the device -- until one seems to "fit, snugly" (differences in IDs can cause a connector with the correct OD to fit but the ID fails to make contact).

The brick is powered on and the device checked for signs of life. Depending on skill level, the user may conduct a cursory evaluation of the device (e.g., note bad pixels on a monitor).

If the device gives the appearance of working, the color code on the selected barrel connector is consulted and a brick from the "*12*V brick supply" having the same color code is selected and dispatched with the device; the octopus remaining in place to check the next device.

If the device appears dead, it moves to a disassembly stage where it can be reduced to recyclable parts.

Some other person processes devices that have been "pre-checked" in this way, looking for true functionality (e.g., perhaps testing ALL inputs on a monitor, verifying connectors are intact -- no "missing BLUE drive", etc.)

Now, put yourself in the position of that (developmentally disabled) "user". You have a color code to remember. WHILE you are looking (elsewhere!) for another device (power supply) that shares the same code. You have to recall that BLUE-GREEN is equivalent to GREEN-BLUE as you search. Yet, rule out each of the other color combinations that you encounter. All the while knowing that you might NOT find the color combination sought! (Yet, you don't want to prematurely abandon your search as you *have* a device that appears functional -- save for the power supply!)

"Devices" take up space so you don't want to retain anything that appears broken (storing broken devices out of doors where space is plentiful and the weather won't affect their NONfunctionality).

And, you don't want to pass devices along for more critical evaluation because there aren't many folks capable of doing that -- especially on a daily basis.

Other use cases: "Tommy, bring over a dozen YELLOW-RED (1TB SATA) disk drives so we can rebuild these computers" "John, use the BLUE-GREEN 'screwdriver' to disassemble these cases" "Valerie, you'll need WHITE-RED power cords for these computers" "Guys, fans go in the BLUE-YELLOW container AFTER you've removed any steel parts"

Remembering RED-GREEN -- and, realizing that GREEN-RED is equivalent while searching -- is likely easier to "do correctly" than remembering YELLOW-BLUE-BLACK-WHITE is equivalent to BLACK-BLUE-YELLOW-WHITE, etc.