In real life, “open mike” terminals are becoming popular, that listen for commands addressed to them and answer questions, perform online actions, or operate compatible smart-home controllers or an ever growing number of IoT devices. In ten years (when my story is set) they may even work as well as they do in the commercials.

But what's a totally different way to perform generic or pre-scripted commands in the home? It doesn't have to be a full language, just handle things like adjusting the lights, setting “modes” (group presets for any number of devices), make the temperature warmer or cooler, and other ubiquitous things in a home. One universal command can be to “listen!” so then you can talk to it, for more complex things.

I envision a lavish home where the characters control the lighting while moving from area to area without having to go to light switches; just automatically do it without breaking stride. It also controls temperature in a shower, operates window blinds, and such.

I recall in on of Asimov’s robot novels he describes controlling things with hand gestures. I don’t know if that’s great, though — how does it monitor your hands?

What other ways might one command a smart home or specific stations therein?

  • $\begingroup$ That is idea generation. But a quick idea: a heads up display.... either a lightweight head mounted display — say for instance that something that has the form factor of plain old glasses can display information — or HUD contact lenses.... coupled with arm & hand tracking of the sort that Microsoft Hololens does. Then you will have virtual controls overlaid on reality... buttons, sliders and whatnot that you can click on using your hands and fingers. The way Hololens does it is simple in its concept (but technically bloody difficult): it looks at your hands and reacts to certain gestures. $\endgroup$
    – MichaelK
    Sep 20, 2017 at 12:48
  • $\begingroup$ Related perhaps: newscientist.com/article/… $\endgroup$
    – Raditz_35
    Sep 20, 2017 at 12:49
  • $\begingroup$ hand gestures are quite feasible provided you have enough cameras everywhere so your hands are hardly ever out of sight to them. But i think diversification could be key here. Switching lights on and off can simply be controlled by the presence of a person. Temperature might be balanced to the clothing of the occupant, or their skin temperature as measured by an IR camera. $\endgroup$
    – Burki
    Sep 20, 2017 at 12:53

4 Answers 4


Neural Interface

The most obvious way - to me - is some sort of neural interface. It doesn't have to be an invasive implant, or some super advanced tech either.

It could be something as simple as a sensor band you slip on your head, and which is capable of detecting certain thought patterns. At that point you could program in a certain thought patterns (make a fist 3 times quickly), and assign automated responses to them.

You can make the tech more elegant by not having physical gestures associated with it at all, and simply having it interpret when you're thinking about turning on the light, etc.

Sensor Band

The other way you could do this is have hand gestures interpreted by a sensor band worn on the wrist/forearm. This technology actually already exists, and was marketed as a computer mouse of sorts.

The band sits very tightly on your forearm, and can actually detect which muscles you're flexing in your hand as you make certain gestures. You can then associate those gestures with commands. It also featured a gyroscope/accelerometer which allowed the users to wave their arms while making certain gestures in order to skip slides, drag windows around, etc.

Unfortunately it's not very precise, and becomes rather uncomfortable to wear (it has to be on very tight).


Some thoughts...

First, I do think that voice control is the way to go, because it's a natural. People like to talk, and they like to be listened to. ;D But maybe there are some other metaphors. Hmm...

Stare recognition -- Let's imagine that there are controls for heat, light, music system, and so on. Each control would be in a different spot, and would use image-recognition to detect when a person was staring right at it. Then the next glance the human makes would be a command. So ... stare pause glance up would turn up the heat, or brighten the lights.

Sign language -- I believe this has already been mentioned; it would be another good option.

Syncopation -- Each appliance would have a different "beat". "Clap...clap...clapclapclap" might be the signature for the shower temperature, f'rex.

And a special bonus feature!

Memory -- each device remembers how you like to set things (lights, temp, etc) under certain conditions. When each device sees you, it puts things just how you like it. You can of course override with standard methods (dials, neural implant, voice). So when you bring Mari-Jo over, your stereo automatically goes into "date mode" and puts on soft music. And when Mari-Jo realizes that the song playing is Karen's favorite song ... well, you're on your own. ;D

  • $\begingroup$ But they don’t like to have everything they say listened in on all the time. That’s the premise. $\endgroup$
    – JDługosz
    Sep 21, 2017 at 16:10
  • $\begingroup$ I remember stare recognition used in SLR cameras and other things but it has problems. I think it can work if combined with something else so “focus on what I’m looking at right now” button is easier than joystick fingering. $\endgroup$
    – JDługosz
    Sep 21, 2017 at 16:13
  • $\begingroup$ @JDługosz my personal preference is voice, but I understand that that's out of scope. My second choice is syncopation, just because it's ... fun. Of course, if you're listening to a Carlos Santana drum solo, your house may explode... ;D $\endgroup$
    – akaioi
    Sep 21, 2017 at 16:26
  • $\begingroup$ that actually makes sense for a spa-shower room: drum your hand on any tile wall to issue commands. E.g. room is currently doing colored rain (lights in the emitters) with an accompanyment of bells; last person there, relaxed on a bench along the wall, wants to change to her favorite “rain at night” program. So she taps on the wall she’s leaning on. $\endgroup$
    – JDługosz
    Sep 21, 2017 at 18:24

What other ways might one command a smart home or specific stations therein?

You (almost) wouldn't need to.

Most of the time, unless overridden by verbal commands, the smart home would watch you and anticipate your needs.

You enter a room, the light comes up (easy already: IR sensor).

You look at the stereo, the stereo turns on and tries your favourite songs. This also is almost here ("OK Google, play Midnight Serenade"). Some intuitive gestures (a hand moving up or down) could then control volume. The stereo would however detect other sound sources and lower the volume if you started talking with someone or received a phone call.

Your own body surface temperature, facial circulation, heart rate, perspiration etc. can be monitored (currently only with contact sensors, but tomorrow...?) to determine whether you're feeling too warm, cold, or okay, and control both temperature and humidity.

You enter the bathroom, put your hands in the sink and the water flows (already exists); or you take off your clothes and enter the shower, same deal. Plus something like a Roomba fetches the clothes to have them washed.

I recall in on of Asimov’s robot novels he describes controlling things with hand gestures. I don’t know if that’s great, though — how does it monitor your hands?

With its eyes, just like a human servant would. Except they're cameras. And can see in the dark.

You could do all this with a neural interface, and if you already had one you probably would. But the interface replaces the need to speak; the need for concentration is the same with or without interface, as long as it's you who has to micromanage the not-so-smart house and decide 'Turn on the lights' and the like.


The ideal wired house would combine many of the current features found in today's "smart" houses with high fidelity body positions sensors, something in effect like a Wii.

Many homes already have devices which can track your movement and do things like turn on the lights when you enter (or turn them off when you leave), send music or tv signals to the devices in the rooms you are in (in effect the show follows you around the house), adjusts room temperatures and so on.

Added to that, most of these set ups have "memories", so if you have "trained" the house to turn on a bright light in one room but dimmer lights in other rooms, that is the default setting whenever you enter the room.

A high fidelity sensor simply allows the user to make finer adjustments. You enter the room and the lights and radio come on, but you can now point at the radio and adjust the volume by raising or lowering your hand. Other fine adjustments could be made depending on how things are wired up, such as opening or closing curtains and window blinds. Temperature could be adjusted by directly measuring biometric information of the person in the room, and indeed other devices could be engaged as well (the biometric sensor senses stress due to elevated body temperature, elevated heart rate and rapid breathing; once it is satisfied there is no medical emergency it gradually lowers the lights, adjusts the room temperature and plays soothing music or white noise to destress the occupant).

The high fidelity sensors would also (aided with a suitably powerful software suite) be able to distinguish between the various people in the house, and use the proper "memory" setting for each individual, with some sort of "averaging" feature when multiple people are in the room.

I would personally find this sort of self contained system more suitable for security conscious people than "open mike" systems. I already have friends complaining that they are being spammed with ads based on conversations held inside the house in private (triggered by key words in the conversation), and some even suggest android and iPhones also now listen to you by default (most peculiar are the people who only have iPhones in house, but not "open mike" systems like Alexa who still see spammed on line advertising based on conversations held inside the house). What you say and do at home should remain at home, not being datamined and sold to the highest bidder.

  • $\begingroup$ Poor people who don’t know how to block ads… I guess they pay for the rest of us too. $\endgroup$
    – JDługosz
    Sep 21, 2017 at 18:28
  • $\begingroup$ Even if you are using adblockers, it does not change the fact that your data is still being collected and sold. Using adblockers simply makes your profile less valuable.... $\endgroup$
    – Thucydides
    Sep 21, 2017 at 21:30
  • $\begingroup$ Maybe can counterattack with a bot that surfs when the machine is idle. Have it average out across people, and provide plausible deniability for any site you want to claim you didn’t visit. $\endgroup$
    – JDługosz
    Sep 22, 2017 at 16:50
  • 1
    $\begingroup$ AdNauseam: adnauseam.io. "As online advertising becomes ever more ubiquitous and unsanctioned, AdNauseam works to complete the cycle by automating Ad clicks universally and blindly on behalf of its users. Built atop uBlock Origin, AdNauseam quietly clicks on every blocked ad, registering a visit on ad networks' databases. As the collected data gathered shows an omnivorous click-stream, user tracking, targeting and surveillance become futile." $\endgroup$
    – Thucydides
    Sep 23, 2017 at 0:52

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .