Development of tiny drones with the ability to move between obstacles and inside buildings with the efficiency of a fly

A research laboratory led by Prof. Beaz Ben Moshe, Head of the Department of Computer Science at Ariel University, develops cheap drones for challenging tasks, with the help of integrating a diverse team from different fields and stages of study

 

In an era where the use of drones is only increasing, it is not surprising that a large number of advanced research laboratories from all over the world are investing a lot of resources in the field of drone research. Today, a drone is used as a replacement for many human roles and this ability is only expanding into different fields.

There are many types of drones in the industry, of different sizes and their hardware world is full of innovations and excellent sensing capabilities. Mainly the quality of the cameras is improving at a fast pace (as in mobile phones). However, the topic of the autonomy of drones is still in its infancy, and with regard to flying inside buildings and between obstacles - the capabilities of autonomous drones are still very far from that of a fly, bee, or wild pigeon. How can we develop a "smart" drone that will be suitable for autonomous missions and with a minimal budget?

This is where the software comes in. Drones today have a variety of sensors, the challenge is to efficiently and in real time analyze the information from the sensors to enable autonomous flight.

For this purpose, the research laboratory gathered K&CG  of the Department of Computer Science from Ariel University. The research laboratory, led by Prof. Ben Moshe, an expert in the world of autonomous vehicles, invests many resources in research and optimization of the world of drones in order to use cheap drones with light hardware for the performance of complicated, challenging tasks that require drones with the most innovative and high-quality hardware.

"I believe, explains Prof. Ben Moshe, in combining researchers from a variety of fields of thought and ranking different degrees, starting with bachelor's, master's and doctoral students, to produce original and diverse research. Therefore, she decided to offer the student audience from the Department of Computer Science and the Faculty of Engineering to participate in a group competition which deals with tiny autonomous drones".

The competition consists of three main stages: an introductory meeting, a qualifying stage and the final stage - each team must take part in all three stages. When in the qualifying stage, the teams will present skills in computer vision* (Computer Vision) in order to be eligible for a drone for the final stage.

In the final stage, an obstacle course, when the teams will develop an efficient algorithm, with the help of which the drone will perform an autonomous flight on a complex course inside a building. The winning team is the one that manages to do as many laps of the track autonomously as possible in a full 5 minute flight.

Each group that demonstrated image processing ability received a miniature drone of sorts TELL which includes a basic case also in terms of hardware, which weighs less than 100 grams. The drone connects to the computer to transfer the data. Each group will develop its own software system. Driving the drone is autonomous meaning without human intervention. The competition is held inside a complex structure and the navigation itself is based on "signs" in a known location (See extension ).

"I believe, adds Prof. Ben Moshe, that making the topic accessible to male and female undergraduate students will intensify interest in the topic and arouse great curiosity in the student audience - who are the next generation of future research. I and the faculty of the department hope that this competition will form the basis for an annual competition on the topic - when future competitions are expected To be open to all interested public - from high school students to graduate students."

First place went to Sa'ar Tobihu, Alon Barak and Eyal Hagai Atudai students in the Department of Computer Science for developing an algorithm that performed the route in the most accurate and efficient manner.

*The research laboratory K&CG  (KINEMATIC & COMPUTATIONAL GEOMETRY LAB), the laboratory for kinematics and computational geometry

Computer Vision* Computer vision is a major branch of computer science, which deals with the automatic processing of images and video, with the aim of extracting and interpreting the visual information contained in them.

More of the topic in Hayadan:

• Development based on the properties of the lotus plant will save the vineyards in Europe
• Learning from the heat of the sea: production of forged materials underwater
• "One accident is enough to cause the drone market to collapse"
• The air is the new internet
• The Naama project conducted an experiment in a national drone network for deliveries, transporting medicines, mapping and more