Advice on using screenless Android phone for IoT device

[Bruce Rankine]

I would like to create an IoT device by buying new, cheap android phones, strip them down and remove the screen, rebox into my own physical box, install a custom ROM without any bloatware (and that will boot without a screen!), and install my android app on the device to do stuff.

An example of a purpose for this could be a GPS tracker for a car. The box would be placed in the car, and record GPS and accelerometer readings, posting these readings back to a central server via the cellular network. (This is just a random example, so don't focus too much on the detail of this, but there are thousands of uses for a IoT board with the sensor, CPU, RAM, storage, and connectivity capabilities of a budget android smartphone)

The reason I want to use existing phones is that they are wonderful, mass produced, cheap devices with a variety of sensors I can use.

The reason I want to use Android is because it is because of the customization ability, and the mature development ecosystem.

To me, it seems an obvious thing to do, but I don't seem to be getting much joy trying to search for examples of this sort of thing (either here, or on the internet in general).

So some questions:

1. General thoughts? (Good idea? Am I missing some fundamental problem?)
2. What are the challenges of running Android without a screen connected?
3. Are there any custom ROMs you know of that specialize in this sort of thing?

(I've seen Google Brilo, but it still seems a bit early yet, and I really like the idea of just using the standard Android SDK to develop the app - and the abundance of help and information that comes with it)

Thanks!

[Crt Mori]

Android without a screen? I think Embedded Linux is a lot more suited to this case. Android SDK for the app without a screen will be quite a fun feat (not really what Android was meant for).

BTW on side note: If you build a dedicated device for your case you will get it a lot cheaper even in smaller quantities, because board will be smaller, less obsolete stuff and CPU does not really need to be that fast, so you save at least 10 USD there.

--
--
unsubscribe: android-porting+unsubscribe@googlegroups.com
website: http://groups.google.com/group/android-porting

---
You received this message because you are subscribed to the Google Groups "android-porting" group.
To unsubscribe from this group and stop receiving emails from it, send an email to android-porting+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Bruce Rankine]

Thanks for the feedback - I didn't know about the Android distribution for the Pi!

Regarding the cost, I'm not so sure. For the likes of a Raspberry PI I would also have to source and buy sensors, flash, battery, charging circuit etc.

As a consumer I can buy a budget smart phone (e.g. Huawei Y3 - I bought one yesterday to experiment with) for $43. Considering it already has flash etc on board, from what I can see this makes it more fully featured and cheaper than the likes of a Raspberrry Pi. Furthermore, I could buy bulk wholesale from a mobile phone manufacturer direct, and even say that I don't need the case of the screen, I'm sure that the $43 could come down quite a lot further.

Trying to design, manufacture, write software for, and support my own board for less than this would be a challenge!

On 4 November 2016 at 13:24, Igor Kalkov <k4l...@gmail.com> wrote:

Android can be a good choice if you want to speed up the development process (Java is easier to learn than C/C++). Android has more easy-to-use libraries and components available. I can understand why it MIGHT be a good idea to work with it.

Nevertheless, Crt Mori is right: its way cheaper to buy a small embedded board. What about Raspberry Pi 3? BTW: there exist an Android distribution for it! (Google for RTAndroid).

[Igor Kalkov]

[Arne-Christian Blystad]

You should be able to boot Android without a screen, it at least works fine on TI and Qualcomm dev boards, though it might require some changes in the kernel. Then you can use Android screencast or similar and have a virtual screen. (Useful during development for navigating to the settings app and similar).

Then write an app that starts during boot, and does what you want.