International Micro Air Vehicle (IMAV) Competition, 2014


The NUS UAV Group from the National University of Singapore (NUS) has participated in the International Micro Air Vehicle (IMAV) Competition, which was held in Netherlands, August, 2014. The competition consists of a single mission that combines both outdoor and indoor mission elements. The focus is put on the following tasks: surveillance, recognition, endurance, and multi-MAV operations.

The team has won the 1st prize in the competion. The 2nd and 3rd places went to Germany and France respectively. Their achievements are also featured in NUS website.

 

 Competition Tasks

 

Overall Mission Scenario

The competition takes place in the village of Oostdorp, a training­village consisting of 30+houses near the real village Harskamp. A major natural disaster has occurred in the region surrounding Oostdorp. The team arrives at the Local Command & Control (LCC) center that has just been set up next to the village. The current situation in the village is largely unknown.A number of critical tasks should be perfomed to ensure that emergency services can be deployed effectively in the coming hours.

The mission includes the following tasks: a) Create a detailed map of the Oostdorp area, indicating which roads are still available and which roads are blocked; b) Perform an initial 'house­by­house' visual scan for any survivors; c) Perform a detailed visual scan of the interior of a building that is already known to contain several wounded; d) Observe one of the buildings that is a potential hazard to the emergency services being deployed.

(Download Completed Competion Rules)

1. Take off

Mission element 1: Launch MAV.


2. Create photomap of village (A)


The purpose of this task is to create a detailed orthomap of the village and use it to present information about the current situation to the emergency response teams.

Mission element 2: Create a photomap.

Mission element 3: Create map of blockages.

3. Scan buildings (B)


The purpose of this task is to do a close visual inspection of the buildings and look for survivors.

Mission element 4: Quick visual inspection: Each building is numbered. Points are awarded for correctly recognizing these numbers.

Mission element 5: Detailed inspection. Some houses can be entered through a door or window. Points are awarded for entering a house and correctly reporting the number of survivors.



4. Inspect building (C)


The purpose of this task is to create a overview of the rooms in a building and their contents. The building is indicated on the map as 'C'.

Mission element 6: Points are awarded for each room visited.

Mission element 7: Points are awarded for each correctly identified and located item in a room.

5. Observe building (D)


The purpose of this task is to observe a building for a longer period. The buildings are indicated on the map with 'D', where the left building with the flat roof can be the 'perching' location, and the right building the actual observation target location. The picture on the right shows part of the roof of the perching location, as well as the side of the house that needs to be observed.

Mission element 8: Landing on the roof of the building with the flat roof. The MAV must remain at this position for at least 10 seconds and has to be able to take off again.

Mission element 9: Observe panel. A sequence of digits will appear on a large panel positioned in a window on the first floor of the west wall of the building with the red roof, during the entire competition slot.

 

6. Precision landing (E)


Mission element 10: Landing. The landing zone is indicated by a large circular sheet, and is for all types of MAV's. Different dimensions apply for fixed wing and VTOL MAV's.

 

 

 


 Competition Results

 

1st prize
National University of Singapore (Singapore) [Score: 683]

Onboard automatic image stitching, onboard number recognition(*), onboard autonomous laser-based room navigation, onboard computer vision based precision roof landing, autonomous takeoffs, autonomous landings (*), onboard computer vision based 7-segment digit recognition (*), autonomous flying WiFi-relay.


2nd prize
Team Dipole (Germany) [Score: 425]

Smallest MAV's of the competition. Very talented FPV flight with many take-off, precision landings, reading house numbers, visiting 18 indoor rooms (several double and not counted), recognizing 16 indoor objects correctly. Geocopter auto-take-off and flight, precision auto landing and partial high resolution ortophoto.


3rd prize
Ecole Nationale de l'Aviation Civile (France) [Score: 189]

Autonomous takeoff, Best overall photomission, all blockades visible, full village high resolution map at 6cm/pixel, autonomous computer vision based 7-segment display reading (**), longest correct observation string, autonomous roof landing (*), reading house numbers with many ARdrones in autonomous flight. Several autonomous landings.

(Download Completed Competion Results)


 Photos

 

The flash player is required to view photos.