Aircraft That Don’t Ask For Directions

Dean Sigler Electric Powerplants, Sustainable Aviation Leave a Comment

During the last two Electric Aircraft Symposia, Sebastian Thrun has shared his visions of future autonomous highways travelled by free-range cars that literally think ahead of the curve and don’t allow themselves to be boxed in – and even more daunting – autonomous helicopters that independently perform maneuvers that stretch the envelope in new directions and dimensions.

His 2009 EAS presentation featured a Defense Advanced Research Projects Agency (DARPA) desert race in which his Stanford University team fielded a Volkswagen Taureg in a 132-mile race through the Mojave.  Although not the ultimate winner, Stanford’s entry completed the course in a time that would have done pride to any human Baja race driver.

More related to daily driving and eventual incorporation into a “smart” air traffic control system, Stanford’s entry in the DARPA Urban Challenge showed what is possible in close-quarter driving.  As Thrun noted, careful measurements from aerial and satellite photographs show huge gaps in what is considered “bumper-to-bumper” traffic, with 80-percent of the road not used even in high-traffic corridors.  It will take smart cars to make up for the average driver’s less efficient reflexes to close these gaps.  Thrun’s reasoning and the attendant hardware/software packages would reduce the need for new roads, save fuel, and make commuting a pleasure while allowing individual freedom of movement.

To cap things, Thrun showed videos of the Stanford Autonomous (model) Helicopter performing incredible maneuvers that required self-learning on the part of the vehicle’s software.  The craft improves over time without human intervention.

At the 2010 EAS, Thrun introduced his compatriot Dr. Nicholas Ray, Associate Professor of Aeronautics and Astronautics at Massachusetts Institute of Technology and head of CSAIL, the MIT Computer Science and Artificial Intelligence Laboratory, or the Robust Robotics Group.

According to their web site, “Our research goals are to build unmanned vehicles that can fly without GPS through unmapped indoor environments, robots that can drive through unmapped cities, and to build social robots that can quickly learn what people want without being annoying or intrusive. Such robots must be able to perform effectively with uncertain and limited knowledge of the world, be easily deployed in new environments and immediately start autonomous operations with no prior information.

“This engineering challenge will require algorithmic advances in decision-theoretic planning, statistical inference, and artificial intelligence. We specifically focus on problems of planning and control in domains with uncertain models, using optimization, statistical estimation and machine learning to learn good plans and policies from experience.”

At the Agra, India 2008 Micro Air Vehicle contest, these tiny vehicles used GPS, Inertial Measurement Unit (IMU) guidance, compasses and a variety of sensors to fulfill their mission of finding hostages on a flat plain.  “Finds” had to be performed within 12 minutes because of the micro endurance that goes with the crafts’ diminutive size.

The 2009 Association for Unmanned Vehicle Systems International Aerial Robotics Competition indoor event in Puerto Rico challenged the micro-copters to find a light emitting diode (LED) while searching through featureless hallways and around various obstructions and distractions.  MIT’s winning entry took little more than five minutes to seal a solid victory.

As this type of autonomous aerial vehicle grows in number (Air Force numbers show far greater acquisition of UAVs than manned aircraft), we must look at integrating their more frequent flights into the overall air traffic control system, already stretching its limits.

To pull all these threads together into a cohernt fabric, the fifth annual Electric Aircraft Symposium will feature Dr. Roy, addressing the issue of autonomous electronic flight contols, and Richard Glassock from Queensland University of Technology, Australia, discussing the Smart Skies project – an automated airspace control system.  This is a culmination of the technologies that will make future flight a reality.  Be sure to attend.

Leave a Reply

Your email address will not be published. Required fields are marked *