Category Archives: Adaptive Technology

Robotic Arms And Snakes

I’ve always loved robotic arms. One of my favorite toys was a Radio Shack Armatron. Which was quite a piece of mechanical engineering. No servos or steppers in it! Instead, it had a continuous running DC motor engaged to a series of clutches, gears, and other mechanical connections. All the “oomph” was in the base. Power was then transferred to the parts of the arm. I learned this, of course, by taking mine apart. You learn so much by taking things apart. I highly recommend it.

Well, most of the robotic arms these days solve the problem, in my opinion, incorrectly. They put motors everywhere. Tiny servos in the “fingers” more servos at each joint. All of which take up a lot of space and weight.

Why incorrectly? Because many seem to ignore millions of years of biological evolution. Do you have motors in your fingers? Are your fingers very muscular? How can your relatively small fingers hold your entire body weight? Because the strength isn’t in the fingers — it’s in the forearms. Tendons do the dirty work. The power is located in another section of the arm. Where it’s okay to be bigger.

So, when working on a project which would require robot arm-like movement, and be small, and flexible, I went back to nature, and groundwork laid by some pretty heavy-duty industrial uses. I went to the snake.

Instead of a series of motors in the arm, there are mounting points for steel cable. The cable feeds down the arm to an array of motors and their controllers. The motors “push” or “pull” the cable to flex the arm. Want it to flex at another point? More wires. and motors

It can get complicated really quick, and control needs some fancy math, but you end up with a very flexible arm which can fit into unusual spaces. For example, large versions of these have been used to access the cores of nuclear power plants. They are also used to inspect and clean the insides of jet engines. Places were human hands can’t reach and traditional arms too clumsy.

This is not to say that more traditional servo driven arms aren’t great. They are used all over the place. But, for some situations, you have to go back to nature and follow the advice of millions of years of development.

It’s early days, and I may very well give up on the snake idea, given its own complexities. Mother Nature is very much “smarter” than me. But, sometimes, you have to attack a problem from a different angle. Or with a snake.

Improving Velcro Buttons

I’ve recently been in touch with a physician who actively works with patients encountering a number of issues. It could be Parkinson’s, or Spinal Injuries, or ALS, or arthritis, or whatever.

One of the mechanical issues he described was the inability for the person to dress themselves, especially when it comes to button shirts.

Latches, like in saris (sarees) are a possibility, but still require a level of fine motor skills, and retrofitting (ie. removing existing, sewing on new).

From my initial research, two directions are being pursued: Velcro and Magnets.

You can currently purchase Velcro closure shirts which look just like regular button shirts (the buttons are false). Limitation: Your style is limited to what the manufacturer produces — you can’t just get “your favorite shirt”.

Now, you can modify your favorite shirt, but that requires sewing skills, or knowing someone who can sew, or paying someone to sew. And each shirt must be custom modified.

There has to be a different way. I’m a big fan of using magnets (where appropriate), but my wife Evelyn brought up an great point — what about when you wash the garment? That shirt is going to get stuck to the metal sides of the washer.

So, I have several design goals: It should not require sewing. It should look natural. It should be easy to modify a favorite shirt — or lots of shirts. It should survive the wash. And, importantly, inexpensive and easy to replicate.

This directed me to Velcro dots, which are inexpensive and available — but often require sewing and/or ironing. This was a good start with the “grabbing” part of the problem. But what about the buttons? And the no-sew?

3D printing to the rescue, with a design which could be injection molded in large quantities.

A prototype I came up with is a two piece solution. The “inside” part slips over the existing button on the shirt. On it is mounted Velcro.

The “outside” part slips through the existing button hole, with a faux button on the outside (on which you could, if you have special buttons, glue a regular button to). It has Velcro mounted toward the body.

Since these are 3D printed (at least initially), button size and color can be adjusted as necessary. I’m thinking of trying some of the buttons in a faux wood filament that I have.

From a caregiver perspective, they are easy to install and easy to remove (if you wish to do so before washing, or if you have limited funds and can only afford one set). No magnets to stick to the washer. No sewing — in fact, no modifications whatsoever to the shirt.

And you can use the shirt you bought at Goodwill — you don’t have to order a $45 shirt from a manufacturer. (No disrespect to Goodwill — nearly all of my daily wear button shirts are from there.)

Since they aren’t a huge bother to remove, I’m also going to try a version with magnets instead of Velcro. This may allow me to reduce the size of the parts.

I’m also in the process of measuring every button I can find, to see what the most common sizes are. The models are currently designed to fit many of the shirts I own.

Me being me, the design model is completely parametric and in OpenSCAD — so adjusting for different sizes isn’t a big deal. Plug in the diameter and thickness and the model adjusts automatically. Something that can be posted on Thingiverse with their Customizer — so the end user wouldn’t need to know OpenSCAD.

Updated Timer Switch For AT Users

So, I’ve been following a thread on the ATMakers FB page about a Switched Bubble Machine. Lots of great info for a maker like me.

The latched and/or timed switches appear to have been around for quite some time. Me being me, I felt the solutions need to be revisited. During my research, two things kept popping up: cost and features.

So, I’ve got a working prototype which has all the features that I’ve seen so far (except machine learning–which is an excellent idea). Jack for input, another for output. Both 3.5mm-1/8″. Adjustment for duration and modes (passthru, timed, latched). User feedback, both visual and audio. Compact. And it runs on 2 AA batteries (and can run on 4, for extended use).

For the users: What features would you like to see in a smarter timed switch?

For the techies, it’s got a QT PY samd21. All parts sourced from Amazon or electronics suppliers (digikey, mouser, allied, etc). All current, and available, production parts.

Feedback is appreciated.

Custom Joycons In The Wild!

To those who’ve been following the adventures of the custom JoyCon style stick: They arrived safe-and-sound. I had a Zoom call with the gamer on Wednesday and they are working for him! He’s lost more strength in his thumbs, so is using the stick with his big toes. We’re now looking at more toe mounted options, along with some thumb-mounted devices which only require movement, not holding a stick. Thanks to all who contributed to get this out the door before Christmas!