There were some challenges on the way to the finishing line.

Not all of the drones will take Lego as food

Even though we’re generally happy with the drones we tried to achieve our goals with, they were not so easy to force to observe small Lego parts (suspicion: small size). While the task is doable, it would probably need to fly very low and very carefully. The most suitable object would be of size of a melon.

Drone + Azure = not always success.

The main project goal was to extract, transfer and load some data from the drone to analyze it for future use in Azure services. Naturally, a connection to Azure is essential for this task. To set it up showed up to be easier said than done. But that remains a task for future studies.

Struggling with the basic drone control

Even though our team succeeded in ordering the drone with the basic tasks, it was still not a streamlined process and required much attention (and time) for each test.

Summed up, it was an interesting challenge, but this time it was too challenging! 🙂

Connecting to the drone has been a small challenge in itself, but with that behind us, we’re moving forward.

Here we are trying to control the Drone using Code, for this purpose we have selected Python as our coding language.

So while flying the drone, It will capture the Video and Images of the objects like Legos.

Optical recognition technology will recognize real life objects and make it possible to store the data. Our mission is not to reinvent optical recognition, but to concentrate on Extract, Transform, Load (ETL) process into the Azure cloud. We want to be able to discover certain objects (lego men) during the drone flight, and perform certain actions if those objects are discovered (eg. send Vipps invoice as a matter to repay for being unable to hide from drone).

Lego Object recognition

As a minimum, we want to be able to detect lego men and be able to run an action if any objects of this type are present. That is a simple task, requires no connection to any form of table/db.

What is the top bar for the scope of this run, is to be able to distinguish between different Lego men, either by face recognition (that is a very high expectation), or just color/object recognition

In the end: we want our drone to be able to collect optical data, transfer it to the Azure, compare with any existing data in the table/database and make a base for a report.
(Diagram coming later)

Drone 1st flight – operated through code

Connecting to the drone has been a small challenge in itself, but with that behind us, we’re moving forward.

Here we are trying to control the Drone using Code, for this purpose we have selected Python as our coding language.

So while flying the drone, It will capture the Video and Images of the objects like Legos.

Connecting to the drone has been a small challenge in itself, but with that behind us, we’re moving forward.
Here’s a historical video of the first flight over beautiful ACDC landscape