This post first appeared on www.advancedwireless.org

Three winners are taking home top honors in the AERPAW Find a Rover (AFAR) challenge, having produced the most accurate localization results in an exercise to track the radio signal from an unmanned ground vehicle (UGV).

The University of North Texas Eagles Team had the closest estimations in live flight tests conducted at AERPAW’s Lake Wheeler Field Laboratory, while the New York University Tandon School of Engineering team had the most accurate results in tests with their software using AERPAW’s virtual development (digital twin) environment. University of Georgia’s Team SunLab rounded out the top three with results that fell in between the other two teams in some tests occurring in both the development environment and at Lake Wheeler.

Participants in the AFAR challenge used a software-defined radio (SDR) mounted on an unmanned aerial vehicle (UAV) to observe signal strength from a UGV-mounted radio. Competitors had the option to adapt the trajectory of the drone to improve speed and accuracy of localization and used varied technical approaches to accomplish the task.

Thanks to AERPAW’s generous sponsors – Galaxy Unmanned Systems, Unmanned Experts, and AnyMile Platform – the winners of the AFAR challenge will take home a total of $8,000 in competition prizes. The top three teams for the fastest (3-minute) location estimate are the Eagles, NYU Wireless, and Team SunLab, while the top three teams for the final (10-minute) location estimate are the Eagles, Team SunLab, and NYU Wireless, respectively.

In one of the main takeaways from the challenge, participants were able to localize the UGV with significantly greater accuracy using the AERPAW digital twin than they were in live field tests, highlighting the significance of experimentation in real-world environments. For each team, measurements were taken at three different locations of the UGV for both the 3-minute mark and the 10-minute mark of flight, resulting in a total of 15 distinct UAV experiments for the five finalist teams.”

Ismail Guvenc, PI and director of AERPAW, says about the competition: “The AFAR student challenge has been a tremendous success with the participation of five finalist teams. The highlight of the competition is that it demonstrated how AERPAW’s digital twin can be used completely remotely by AERPAW’s users to develop AI/ML-aided solutions at the intersection of advanced wireless systems and autonomous UAVs. Once fully developed in our virtualized environment at the own pace of our experimenters, those solutions can be seamlessly transferred to our real-world testbed for final evaluations.”

Mihail Sichitiu, Co-PI of AERPAW, commented: “We would like to congratulate all the AFAR finalist teams who participated in the competition, but particularly the Eagles team with its first-place finish in both the fast-estimate and final-estimate categories. We would also like to thank our AERPAW operations personnel who worked tirelessly to finalize the competition, especially to Ozgur Ozdemir, Thomas Zajkowski, and Anil Gurses, and to our three sponsor companies that are providing competition awards.”

The AFAR challenge is the first in a series of new drone competitions AERPAW will host. For more details on the competition, please visit https://aerpaw.org/aerpaw-afar-challenge/