Create a free Manufacturing.net account to continue

Designing An Autonomous Flying Robot

For the past 22 years, the International Aerial Robotics Competition (IARC) has been held each year by the Association for Unmanned Vehicle Systems International (AVUSI). Teams are tasked to design and build aerial robots to complete specific missions. This year’s mission: Design and build a flying robot that is entirely autonomous.

For the past 22 years, the International Aerial Robotics Competition (IARC) has been held each year by the Association for Unmanned Vehicle Systems International (AVUSI). Teams are tasked to design and build aerial robots to complete specific missions. This year’s mission: Design and build a flying robot that is entirely autonomous.

"Until the mission is successful, it repeats year after year," explains Danny Ellis, founder and president of the University of Michigan team, Michigan Autonomous Aerial Vehicles (MAAV). "So far there have only been five successful missions, and this is the third year of the sixth mission."

Ellis founded MAAV back in the fall of 2009 as part of a senior design project with fellow classmates. They entered the IARC as a way to gain funding and sponsorship for their aerial robotic designs. "For this year’s mission, all we are allowed to do is hit start, so the robot can perform the task on its own, and then hit the kill switch if we need to," says Ellis.

The goals for this year’s mission include the robot autonomously navigating through an unknown building, finding a target room that possesses a flash drive; retrieving that flash drive, and dropping off the decoy, in less than 10 minutes. Teams gain extra points if their vehicle can recognize Arabic signs, which provide navigational instructions.

The MAAV team has been improving their design for this particular mission over the past two years.

I’ve been working on CAD software ever since I was a sophomore in high school," says Ellis. "When I was a freshman in college, I really wanted an input way to manipulate my model that wasn’t using a mouse, and that’s when I stumbled upon 3Dconnexion."

Ellis introduced the Space Navigator 3D mouse from 3Dconnexion to the student shop. While working on the quadrotor for the IARC, the students noticed how the 3D mouse allowed more opportunities for movement compared to a standard joystick. "We now use the mouse for controlling the flight of our vehicle," says Ellis. "Although we are restricted from using it during this year’s mission, it is extremely beneficial during our test flights, because it makes flying more intuitive." 

3Dconnexion provided all the source code needed to incorporate the mouse as the control system for the aerial robot. According to Ellis, it was very easy to interface, and the integration took less than a day.

The mouse was also beneficial during the re-design process as MAAV worked to incorporate the necessary components while staying under the 1.5 kg mass limit. The vehicle design, and many subsequent redesigns, started out as CAD models. Instead of creating multiple prototype iterations, the team was able to go straight from the CAD file into production. The mouse also helped Ellis to train new team members on how to use CAD software. "The 3D SpaceBall makes it easy to use and navigate through the CATIA software program we use during the design process," explains Ellis. With twelve programmable buttons and an optical/non-wearing measuring system, the 3D SpaceBall creates a comfortable and efficient work environment by eliminating repetitive steps that put unnecessary stress on users’ mouse hand.

During the design process, the MAAV team has faced a variety of challenges that forced them to constantly improve their design. "We’re on the fourth iteration, and we have gone through three full re-designs, but that’s engineering," says Ellis.

MAAV has the quadrotor flying manually, but the main challenge the team currently faces is autonomous flight. "We have two on-board laser scanners that scan the entire environment and return all the points that they see," says Ellis. "Our greatest difficulty is getting the vehicle to learn and respond to its environment on its own, and making its own decisions within that environment."

"When it comes to controlling the vehicle, the 3D mouse produces smoother movements for the pilot," says Ellis. "It also allows us to handle multiple types of input." All of the buttons on the mouse have been programmed to control different things, such as start up, kill, and incorporating some kind of a disturbance to the vehicle. "That way we don’t have to go up to the robot and hit it ourselves," explains Ellis. "We can push a button, create a disturbance, and observe how the vehicle reacts to it."

 

According to Ellis, the 3D mouse is the smoothest device that he has used. The shortcut keys decrease the amount of time he uses the keyboard, and he appreciates the ability to lock and unlock each access of the short-cut keys. "Being able to lock the app keys has been very beneficial, because it allows us to change certain movements of the robots without having to reprogram anything."

 

Ellis and the MAAV team enjoy the IARC and similar competitions, because they see them as segues that help gain support from both corporations and the university to continue the team’s research. Ellis also claims that the IARC helps engineers on the team maintain a competitive edge that isn’t found during their class and career schedules.