RUASRT’s IMAV Outdoor Entry

Matt and Ashim preparing the flying-wing mapper

The International Micro Air Vehicles (IMAV) Competition and Conference has both an indoor and outdoor event. RUASRT took it upon themselves to enter both. For a general overview of how the 9th annual IMAV went, see this page. The outdoor event took place the day after the conference, at an airfield just outside Toulouse. It had several challenges associated with it, the main one being a search-and-rescue style event. Points were awarded for automatic take-off, precision (automatic) landings, mapping the required area, classifying and identifying points of interest on that map, completing as many laps of a designated area as possible within the time-frame, and a cooperative carrying challenge.

Ashim hand-launching the pylon racer

 

RUASRT had two fixed-wing craft to attempt as many challenge elements as possible. This included a pylon racer for the endurance-challenge around two pylons, capable of speeds up to 150 km/h, and a blended-wing for aerial mapping. The competition focuses on high levels of autonomy so more points were awarded when elements were completed with little-to-no manual input from the team.

Matt, Ashim and Ethan planning their test flight

Our pylon racer was capable of completing many laps around the designated points autonomously (and did so perfectly on the practice day), but on the competition day struggled to turn within the restrictive geo-fence.

Ashim and Ethan on the practise day

Our mapping vehicle successfully flew over the entire area required for mapping, during which time photographs are automatically triggered at a desired interval. Once the MAV landed and the photographs were automatically transferred for photo-stitching and target identification, it was discovered that only a portion of the images expected were present. As a result our image stitching software was unable to find enough matches to create the required map. This also meant we did not get a chance to demonstrate our MATLAB script which automatically searched for brightly coloured objects and classified them by size and shape.

Automatically generated map of a test field

The outdoor team was led by Dave Tennent, Dr Matt Marino, Ashim Panta, John Bueker, and Ethan Moyle. The team is very proud of what they were able to achieve, despite the logistical challenge of getting delicate fixed-wings all the way to France. On their return, they plan to invest more time into their mapping MAV to make the autonomy aspect more robust.

Abdulghani, Matt, and Ashim on the competition day

Well done!

 

 

Bookmark the permalink.

Comments are closed