Incorporating a Drone in your Home Automation setup

Hi all,
I’ve started a thread over in the Smartthings forums and I’m wondering if Navio would be a good option for a project like the below. Smartthings (ST) is a home automation platform.

I’ve already got a Roomba hooked up to ST, now I’m looking to expand my robot fleet. Regardless of if the technology is there today, what would incorporating a drone into your HA setup look like?

I suppose the first use case would be for security. Perhaps the drone could fly to where motion is detected when you’re in away mode. It would flash lights and broadcast an audible warning to leave the house. I’d imagine that would be terrifying to any intruders. An accompanying camera would go to where the trouble is, not wait for it to pass by a static security camera. Taking it a step further, LAZERS could get intruders out of your house in a hurry too!

How about for general fetching? Take this to the laundry room. Grab the babies bottle. Put these dirty dishes in the sink.

Another use case might be for cleaning. Dusting hard to reach places. Cleaning tall windows. The airborne cousin of the Roomba…

What would the drone need to do?
Probably need to be a micro drone to navigate mostly indoors
Full autopilot and obstacle avoidance
Camera, lights, etc.
Something to grip things or a platform to put things on
Facial recognition or some other sort of presence detection
You’d need the ability to program flight paths or at least upload the layout of your home
You’ll basically need to teach it what things are in your house. Sink, Washer, etc.
A homebase platform that allows for the drone to land and charge itself
A way to hook into your existing HA setup.

I’m not trying to be a fat human from Wall-E. Just future gazing a bit. So what needs to happen to make this a reality? And can this be done now? I wonder if a homemade drone with a Pi could knock off a few of these things?

This is the conversion on the ST forums:

1 Like

Number 1 question . How technically incline are you with electronics ? And are you looking to make this drone yourself.

1 Like

I do home automation for people with special needs for living. I’d definitely be interested in a discussion about this. We have our own “Internet of things base station”, but I’m constantly experimenting with things like smart things.

1 Like

Technically inclined yes. Have I built any electronics and can I code, not really. I suppose I’d be looking to build this myself, but I think this is first a discussion to see if something like this can be done in the home.

For starters, I’ve only seen larger drones built with Navio. Given this implementation would be indoors, I’m wondering how much miniaturization would need to occur. What’s the smallest build we’ve seen with Navio? What about hardware? Can we incorporate lights, camera, sensors etc? There would have to be a tradeoff in size I imagine. To carry any objects, you’re going to need some lifting power right?

Then, the question is can things like presence detection or facial recognition, obstacle avoidance, dynamic flight path be done with current software? You’d also need a way to hook into your home automation platform. Whether that’s cloud, zigbee, zwave, ble, etc…

Finally, I believe it’s important to have the drone able to land and recharge itself. It’s called home automation for a reason. I don’t want to have to charge the thing every few days.

Are there any other use cases in the home that I’m not thinking of?

@mrdoornbos What’s an IoT base station? A hub you’ve built?

The smallest builds we’ve seen are 250mm. Navio fits pretty well on FPV250 frame, flight time was around ~12-14 minutes without the payload.

to me your drone ‘tasks’ dont seem to be well suited for a drone. More like other robotic machines. At this point in time open source drone code is quite limited out of the box. You need to spend a huge amount of time to add features to it. You can very easily build a drone that will fly indoors well but automated charging and obstacle avoidance is quite the work in progress. Your biggest challenge is flght time and even if you manage to recharge automatically you are still limited to battery cycles for LiPo… I see sooo many challenges but the technology is def there but not available to you. You have to do alot of designing. I would say you have a base HW and SW solution but nothing readily available to integrate.

Thanks @lockheedtristar. I think you may have discovered the 500lb gorilla in the room. The truth is, a drone is probably not the best robot for the home as it is somewhat limited. I do think there is some utility there though. And it could be a fun project. I did sort of assume this is more a 5-10 years down the line thought experiment. Having experience with the rate of innovation, do you think that’s an accurate timeframe?

As @mikhail.avkhimenia states, if we can get 12-14 minutes of flight time, I think that would cover most of these tasks. How long does it take to shoo away an intruder? :wink:

Btw, looking at examples of the 250mm builds, I think that is a good size for in-home use. Not too big, but large enough to support the things we might need. So… that was easy.

Just so I can get it out… here’s how imagine this might work:

You pre-program the layout of your home. Different flight plans to get to different rooms. Flight plan to clean the fan, to dust the tall blinds etc. The HA system would invoke the flight plan needed for the situation.

Then once you get to the end of your flight plan, a new operation would take over.

For example, if you are in security mode, your drone flies to the kitchen, where the motion was detected. Once it gets there a new “Acquire” operation kicks in where it looks for the intruder’s face using some sort of facial recognition or presence detection. Once detected a new “Deter” operation kicks in and the flame thrower comes out.

Or more simply, the drone gets to the blinds and the “Dust” operation kicks in.

Once the battery starts to wane, it would follow the flight path home, acquire it’s base station and land, where wireless charging starts. (This step is skipped in security mode. That’s a one-way mission to protect the house!)

I completely understand the challenges, but I think it’s fun to think about.

ok yes and i did feel like i was discovering the gorilla but what i did not know is; how much assumption there was…

also for a ‘5-10’ year down the road you better start working on a project where your drone carries you from your house to another location. that is very possible in the very near future.

I thought they had pneumatic tubes for that.

Flame throwers and lazers. You have a wild imagination good sir.

I think our attention should stay on non-weaponized applications.

You know more constructive applications like drone dusting apparatus, lol.

Ha! This is a tough crowd I can see.

I’m just painting a picture. A security drone would really just flash some lights, blast an audible warning and look intimidating.


I am sorry if i sounded pessimistic. MY view is this. I have experience in programming and worked as RF Engineer for 6 years now with a telecom giant. With my experience building a drone was reletively straight FW but not easy. However i had high expectations and wanted reliability and to achieve that it took me quite a bit of time/research and frustration. Now when i think about your applications all i see is well ALOT of work and experimentation. You kbow how difficult is to even manage mechanical vibration as noise on these things…its tough. As well as magnetic interference and barometer placement and accuracy, frame design. Looking it from my personal experience when i read your post i was just not seeing it possible without having an entire team of engineers work on that for years.

I would call this group highly pragmatic :wink:

But worry not! When you have “swiffer-quad” v1.0 on the bench and can figure out how to trim out the weight from the duster sticking out in front of your 250mm quad, we will be here to set you straight :smile:

1 Like