The Challenge 

Drones have become very popular in the last years as toys and for photography. However they could also be very useful for industrial applications. But currently their adoption in industrial fields is lacking behind. This is because every drone needs a pilot to operate. This pilot is expensive and can’t work 24/7. This makes it for example impossible for scheduled on location inspections. For example a farmer cant automatically get a report on his crops every day. Or a large manufacturing plant cant automatically inspect its machines every few hours. Besides inspections, drones could also be used to automatically monitor a specific area like a harbor for suspicious behavior. This can help security teams find criminals easier.

The Solution

In order to solve these problems we are going to remove the pilot to make drones autonomous. But this is not a easy task. What we do different compared to everyone else is that we use a unique end-2-end Machine Learning based approach. Our model is trained on real life human flying behavior making it fly just as humans. This will make our system more reliable which is important in industrial applications.

The Mission

Our mission is to deliver an affordable yet industrial-level solution for companies seeking an automated way to perform tasks like inspections or deliveries in a large-scale environment, where humans are costly or can endanger the success of the whole operation. The system developed at MRR offers fully autonomous navigation between GPS locations while avoiding obstacles on its path without human intervention.

This project is being coached by