The Doll Forum

We are all familiar w/t race for AI and an interactive head. Fantastic, but it all cost a lot of $.
I was thinking it would still be nice to have a doll head that would have blinking eye lids and a mouth that lightly moved at random. This would be used in conjunction w/a Bluetooth speaker in the dolls head that could play a moaning tract or other audio such as ASMR whispering, humming, etc. Wouldn't really matter if the lips where in sinc to the audio that much. Would still be a nice visual stimuli for a lifeless face when the lights are down low.

You know TDF's really slow when you have to bump your own post.

Ok, wipe the lube off your hand and hit the keys. You know who you are!!

I totally agree with that. Moving, blinking eyes and mouth movements, even if seemingly random, would be amazing and add a lot more realism. Could just be a battery powered motor with gears, no need for any robotics or AI whatsoever.

Actually I was thinking the same, my imagination tells me that an exoskeleton could also add move to the body and head.

Do I want to spend 7000 on the new DS robot or blow 1000 on a DIY system to make my doll move???

The latest in AI modeling is facial and hand pose estimation.

Just saying.

I remember someone here mentioned about Flexinol actuator. A quick google search brings me here:

https://www.robotshop.com/en/flexinol-en.html

This kind of actuator wire might be the key to your concoction? Like connecting the wire to a tiny bluetooth speaker?

philpw99 wrote:I remember someone here mentioned about Flexinol actuator. A quick google search brings me here:

https://www.robotshop.com/en/flexinol-en.html

This kind of actuator wire might be the key to your concoction? Like connecting the wire to a tiny bluetooth speaker?

That would probably work really well, and easily for any up / down motion (eyes and mouth), or side to side (eyes). The key would be how you regulate the electrical current and the correct intervals. Its probably really easy for anyone who knows about electronics (not me!). Syncing it to sound would be more complicated.

Thanks for the interest.
I researched the Flexinol wire. Common name is Nitinol which is a mix of Titanium & Nickel. Basically dead wire unless heated either direct thermal or electrically. It seems to have a rather slow operation time due to the heat and cool cycle needed for it to operate.

I figure there are 3 actions that can go into mouth movements for this study.
Jaw action that would require a hinged jaw skull.
Lip action and maybe a little tongue wiggle.
It may be possible to tie a 3 into one mechanism through linkage.

actually doll companies need to get more into the animatronic heads as far as selling them , if they are reasonably priced i think they would sell like hotcakes especially if blue tooth equipped and sound / mouth activated .

IMHO:

The current animatronic heads have a whole bunch of servos, gears, and other actuators just to do some simple movements, including blinking.

I'm a proponent of pneumatic soft robotics for a reason: Those servos, etc., are power hungry, and that increases the weight and complexity of the simplest of movements. It also decreases the reliabilty of the system.

However, positive pressure and vacuum pressure soft robotic muscles could do away with all those gears and servos and use the doll's own 'flesh' as the soft robotics muscles.

What's the power source? A single paint ball 68/4500 tank would be a great start.

So, this would require that the doll's head would mostly need to be 3D printed.

That's the only way to embed those kinds of 'muscles'.

the biggest problem with motorized faces is the skin needs to be made thinner so there is less resistance or very strong motors used, the bigger the motor probably the more noise it will make though . that's why disney characters used what was called frubber very lightweight easy to move with jaw mechanisms but some of that doesn't look very realistic a and hard telling how it feels , probably not lifelike .

I do have a source of TPE pellets.

What I haven't done is test out the different grades of pellets in a heated auger extruder.

How can I say this?

We have to build our future.

Pneumatic soft robotics looks promising. One problem is tank refilling. You need to have a paintball or scuba shop close or buy a 2K+ compressor that can fill them. Batteries I can recharge at home.

Most animatronics use a lot of foam and thin latex skin.

I get the issue.

I've looked at Paint ball tank compressors.

IMHO:

It's do-able, cost wize, etc. .

Go to the very long reply I wrote

viewtopic.php?f=87&t=65773&start=15

I've thought about the facial movements and tongue action. If you had a hollow tongue with multiple chambers that expand at different rates/directions, filling each chamber would allow for a pretty realistic movement, but that kind of modelling and calculations are out of my forte.

I can AutoCAD like a boss, so if I had the program, a 3D printer and the Electronics, the coding wouldn't be terrible. I'd be installing some pneumatic actuators for jaw movements, a mechanism for tongue movement and pressure sensors in the lips and mouth for feedback...i have that knock off first doll in storage...I'll keep you guys posted once I decide how much it could cost me. Peristaltic pumps exist for small electronic applications that work well for feeding lubricant. It isn't out of the realm of do it yourself