ICMCIS is again running a data challenge, releasing a dataset for interested participants to develop machine learning based solutions. This year’s challenge is to identify and track unmanned airborne systems (UAS) or drones. Datasets covering a range of scenarios, sensor types and drones are available on Kaggle for the ICMCIS drone detection challenge. Further information is provided here.

The military scenario to this challenge is to improve capabilities to protect people and equipment against the threat of misuse of small (Class I) UAS such as hobby drones. The majority of the solutions developed to counter such UASs so far use a mix of sensors to detect and track drones entering a protected flight zone. Typical sensors are radar or radio direction finding, data from both types of sensor are included in the dataset.

Results of this challenge will be presented and discussed in a special session of ICMCIS. All participants in this data challenge are invited to take part in the special session. The conference will also feature a keynote presentation on the challenges of UAS by Liisa Janssens, co-author of A Comprehensive Approach to Countering Unmanned Aircraft Systems.