Learn, Teach, Invent, Make! with STEMCA Inventor

A project log for STEMCA Inventor - STEM Education and DIY Amplified

The most advanced, affordable and approachable (easy to use) STEM education and DIY platform with focus on robotics and automation/IoT.

Marius SlavescuMarius Slavescu 12/02/2015 at 04:574 Comments

I'm launching a new initiative and invite everyone to innovate for good, and benefit from their creations:

Learn, Teach, Invent, Make! with STEMCA Inventor platform

At the same time, although still a bit rough around the edges, but pretty functional, I'm announcing the first STEMCA App Inventor beta, with lots of interesting features.

To try it out please go and subscribe to our news letter on then head out to and login using your e-mail.

I'll publish soon the companion app in Google Play, for now you can use the APK direct link

I'll publish a few sample apps with the new Computer Vision component in the next posts.

All MIT App Inventor 2 tutorials should work in STEMCA App Inventor.

You can find a very good introductory course for both learners and teachers here:

Educational App Development Toolkit for Teachers and Learners

STEMCA Robotics components learning material will come soon, through videos and sample apps.

On the hardware side, soon we will take pre-orders for STEMCA Arduino brick (that I showcased in previous posts), this is the easiest way to get started with STEMCA Inventor for robotics, automation or any other STEM related projects, very easy to quickly prototype, very affordable to expand.

SPARK (STEMCA Personal Assistant Robot Kit), RC Coller hack and RC Car hack are all using STEMCA Arduino brick to control the motors and read the sensors.

Lego Mindstorm EV3 and NXT can also be used with STEMCA App Inventor.

I wish you Happy Inventing!


Marius Slavescu wrote 12/02/2015 at 06:45 point

This one is more advanced and it would make a car robot to follow a ball (colour blob) using VideoCamera computer vision algorithms:

As before you will need to install OpenCV Manager 3.0 from Google Play before running the app.

All the .aia files needs to be imported in STEMCA App Inventor then using the STEMCA App Inventor companion (sai.apk) you can connect and run the app.

The car can be built using STEMCA Arduino brick or Lego Mindstorms EV3 (the app will need some very small changes to switch to EV3).

  Are you sure? yes | no

Marius Slavescu wrote 12/02/2015 at 06:15 point

This is another example on how to use the computer vision features of VideoCamera component:

Click in the video on a coloured object then click "Get selected color bounds" button, then play with the sliders and see how it affects the colour blob detection.

I'll add soon support to show the live processed image also, so it would easier to see what pixels will be filtered out/in.

If you have ideas or like to have other features integrated please post them here or send me by e-mail at

Basically anything that OpenCV 3.0 for Android provides could be integrated, but soon I'll look into integrating dlib library algorithms also.

  Are you sure? yes | no

Marius Slavescu wrote 12/02/2015 at 06:03 point

This one is to assist you driving, please use it with care! It shows how to use face detection.

You will need to install OpenCV Manager 3.0 from Google Play, and also (temporary workaround, I'll fix it soon) to download this file:

and place it here on your Android phone or tablet before running the app (through companion):


  Are you sure? yes | no

Marius Slavescu wrote 12/02/2015 at 05:54 point

If you have a Lego Mindstorms EV3 try this simple test app, try to control it through whistling :-)

  Are you sure? yes | no