The ICUAS UAV Competition is here again!
This year, we are going from one to multiple UAVs! Coordination, planning, detection and a couple of Crazyflies await new challengers! To register for the competition, use the following link:
https://forms.gle/mk2X5vgsXGujEEYZA
As always, rulebook, code and simulation setup is available on Github:
https://github.com/larics/icuas25_competition
To contact the competition organizing crew, please use Issues/Discussions on the competition repo.
Important dates (subject to change):
January 25th, 2025: Initial submission – team registration closed
February 20th, 2025: Debug submission
February 28th, 2025: Deadline to upload solutions
Top teams will be invited to further develop and showcase their solution at the Conference!
Additional Information
In this year’s edition of the UAV Competition, the teams need to deploy a team of UAVs in an urban environment to locate and identify threats. The UAVs deploy from the base and need to find and identify several targets in a known environment. Since some of the threats may interfere with communication links between agents, the team is required to keep constant communication between the base and all agents in the system (as shown in Fig. 1.).
The team of UAVs is required to locate and identify an unknown number of targets and report the location of the threat to the base. While searching, the battery of each UAV is draining and each UAV can go back to base to recharge, but the system needs to remain connected even with one or more UAVs charging. The UAVs to be used are Bitcraze Crazyflies, running through SITL paradigm to facilitate easier transfer to the finals in the arena at ICUAS’25.
The fourth edition of the ICUAS UAV Competition has drawn interest of 26 teams from India, Hong Kong, China, Brazil, USA, Mexico, Armenia, Saudi Arabia, Pakistan, Czechia, Poland, Italy, Croatia, and South Korea. Following the January 29th deadline for submitting a proof-of-concept solution, 15 teams have committed to the simulation stage of the competition, which ends on February 28th, 2025. Alongside the teams and their UAVs, the LARICS group at the University of Zagreb is working on preparing evaluation scenarios and point scoring scheme that will be used to decide the finalist teams. As is now a long-standing ICUAS tradition, top teams from the simulation stage will be invited to Charlotte to showcase their work in the arena at the Conference. While the competitors and the University of Zagreb crew are busy with the simulation stage of the competition, the group led by Artur Wolek at the University of North Carolina at Charlotte is already working on the competition arena at the conference venue.
ICUAS’25 UAV Competition Simulation Stage Overview and Results
ICUAS’25 UAV Competition Simulation Stage
In this year’s UAV Competition, the task is to deploy a team of UAVs in an urban environment, with the objective of locating and identifying threats. UAVs deploy from a base location, and they must find and identify several targets in a known environment. Some of the threats may interfere with the communication links among agents; therefore, the UAV team is required to keep constant communication between the base and all agents in the team.
Proposed solutions are evaluated in two world scenarios. One world is shared with the competing teams during the simulation phase; the second world is not seen by the teams. However, both world models were made available to the competing teams at runtime, as Octomaps. Moreover, within each world model, multiple ArUco tags were scattered on vertical surfaces within the world, as shown in the below Figure. Each competing team must use a team of Crazyflie UAVs.
The team of UAVs that is deployed must find as many tags as possible within a 30-minute time interval. However, the Crazyflie battery lasts less than 30 minutes! As such, competing teams may consider and include tasks like returning to base to re-charge one or more, or even all Crazyflie UAVs, that is, the whole team of UAVs.
The UAV Competition Organizers developed evaluation software, which tracks the position of Crazyflies checking whether they crash into/with obstacles, fly within flying area set bounds, thus, rewarding or penalizing flying behavior. The developed software was also capable of tracking the battery state of each Crazyflie, reporting if battery level is below a pre-defined threshold, which is also a penalized behavior. Overall, the evaluation software constantly monitors the connectivity of the underlying communication graph by checking the link validity and by calculating the Fiedler value of the graph, reporting each instance in which the graph becomes disconnected. The Figure below depicts a valid configuration of a Crazyflie team.
Valid configuration of Crazyflies in the simulation world. All UAVs fly within the bounds of the flying area, the communication graph is connected, and the battery state of all UAVs is within a pre-specified range.
Teams were awarded points for finding ArUco markers (70% of the total available points) and for maintaining connectivity (30% of the total available points). Points were deducted for each observed penalized behaviors (crashing, flying out of bounds, low battery state).
Finalist Teams
Following a rigorous evaluation procedure, the following teams emerged as top contenders, thus qualifying for the finals, which will take place during the Conference.
- 1st Place: AIRo Lab, The Hong Kong Polytechnic University, China
Team Advisor: Prof. Chih-Yung Wen
Team Leader: Zheng Tan
Members: Li-Yu Lo, Yifei Zhang, Yuzhou Li, Wenyu Yang - 2nd Place: Center for Scientific Innovation and Education – CSIE, Armenia
Team Advisor: Prof. Astghik Hakobyan, National Polytechnic University of Armenia
Team leader: Gor Arzanyan, American University of Armenia
Members: Nane Hakhverdyan, Yerevan State University
Anna Manucharyan, American University of Armenia
Artak Mnatsakanyan, National Polytechnic University of Armenia
Rafik Simonyan, National Polytechnic University of Armenia - 3rd Place: Aerial Robotics IITK, Indian Institute of Technology, Kanpur, India
Team advisor: Prof. Ketan Rajawat
Team leader: Pulak Gautam
Members: Varun Sappa, Vihaan Sapra, Akshat Jain, Shvetang Rao, Ayyappan Atulya Sundaram, Shruti Dalvi, Aman Singh Gill, Anmoldeep Singh Dhillon, Sanskar Yaduka - 4th Place: AGH AVADER, AGH University of Krakow, Poland
Team advisor: Dr. Tomasz Kryjak, AGH University of Krakow
Team leader: Remigiusz Mietła
Members: Hubert Szolc, Mateusz Wąsala, Mateusz Gołąbek, Tymoteusz Domagała, Jan Jagodziński, Kacper Iwicki, Kamil Jędrzejko - 5th Place: KNU ARRF, Kyungpook National University (KNU), South Korea
Team advisor: Prof. Kyuman Lee
Team leader: Joohyuk Lee
Members: Hojun Lee, Jeonghoon Song, Mohomad Aqeel Abdhul Rahuman, Seong-jin Oh, Yonggyun Moon, Kangmin Kim
The UAV Competition Chair and Organizers offer their congratulations to the five finalist teams, but also to the teams that did not qualify for the finals, as they have demonstrated great effort and have shown promising results during the simulation stage. The teams that did not qualify for the finals are: Kaizen (Delhi Technological University, Jaypee Institute of Technology, India), iHunter (Prince Sultan University, Saudi Arabia), ghostdrones (Federal University of Rio de Janeiro, Brazil) and AeroMIT (Manipal Institute of Technology, India).
Competition Winner
Aerial Robotics IITK, Indian Institute of Technology, Kanpur, India
Team advisor: Prof. Ketan Rajawat
Team leader: Pulak Gautam
Members: Varun Sappa, Vihaan Sapra, Akshat Jain, Shvetang Rao, Ayyappan Atulya Sundaram, Shruti Dalvi, Aman Singh Gill, Anmoldeep Singh Dhillon, Sanskar Yaduka
Competition Sponsors
The UAV Competition Chair, and the Organizing Committee members extend their sincere gratitude towards the two Platinum Sponsors whose equipment will enable and allow for the deployment of the teams’ solutions in the competition arena during ICUAS 2025. Bitcraze supplies a fleet of Crazyflies, spare parts, their Loco motion tracking system, and NaturalPoint installs their Optitrack motion capture system. Each team will fly Crazyflies and will showcase their approach to solving the challenging problem posed.
Prepared by Frano Petric