Deploying Decentralized GNSS-Independent Swarm Drones in Urban Environments: a quick primer on Technological Challenges and Frontiers

The realm of robotics has witnessed a paradigm shift with the advent of swarm robotics. These systems are characterized by their decentralized nature and the ability to perform tasks without relying on a Global Navigation Satellite System (GNSS). Their potential in search missions is profound, but so are the challenges. We examine the technology frontiers and challenges by analyzing select references on topics ranging from swarm behavior, communication, and optimization to specific applications in search missions.

Dave

Published in

d*classified

6 min readOct 10, 2023

RSAF-DSTA Drone Swarm being tested at EX Forging Sabre 2023

1. Swarm Behavior and Decentralized Control:

The nature of decentralized control and inherent swarm behaviors in drones is complex, yet fundamental to their operation. Key findings from selected works showcase:

Aerial Swarm Interactions: Manoni et al., 2022 underscored the importance of adaptive arbitration in interactions, enabling drones to autonomously adjust their behaviors based on instantaneous aerial dynamics.
Navigating the Unknown: The capability to employ goal-directed strategies, especially when traversing uncharted urban terrains, has been emphasized by Wang et al., 2022 and numerous researchers from CMU Airlab.
Landing Dynamics: The intricacy of self-sustaining multi-agent systems capable of landing on dynamic platforms points towards an enhanced control paradigm as discussed by Gupta et al., 2022.
Evading Collisions: Vaidis & Otis, 2021 spotlighted the role of nearest-neighbor algorithms in collision avoidance, crucial amidst the dense urban air traffic.

2. Communication and Connectivity:

Communication remains the linchpin of decentralized systems. Urban environments, replete with their challenges, demand novel communication paradigms.

Navigating the Urban Communication Maze: With skyscrapers and other structures often obstructing signals, ensuring constant communication is challenging, a hurdle highlighted by Ashush et al., 2023.
Sensor-Communication Convergence: A harmonized system that converges both communication and sensors is pivotal for urban drone navigation, a notion backed by Perera et al., 2022.
Network Scalability and Integrity: As the number of drones in a swarm grows, maintaining the integrity and scalability of communication becomes imperative, as discussed by Asaamoning et al., 2021.

3. Optimization and Coordination:

The multifarious urban landscape mandates sophisticated optimization and coordination techniques for efficient mission execution.

Vision-based Navigation: Tong et al., 2023 detailed the significance of multi-UAV systems leveraging collaborative vision positioning, enhancing navigational accuracy.
Graph Neural Networks in Swarms: Wang et al., 2021’s work illuminated the potency of graph neural networks in identifying and optimizing communication and task nodes within a drone swarm.
Machine Learning-Driven Drones: ML applications for drones, as explored by CMU Airlab (many good ideas here), Trihinas et al., 2021, hold the promise of enhancing their efficiency and adaptability in urban missions.
Neuromorphic sensors: Neuromorphic sensors, inspired by the human nervous system, offer a novel approach to data interpretation in real-time. Their energy efficiency, derived from event-driven activation, extends drone battery life. With rapid data processing capabilities and reduced data generation, they promise enhanced performance for drones, though integration challenges due to their novelty persist.

Discussion and Analysis:

The synthesis of the aforementioned literature paints a nuanced picture of the GNSS-independent swarm drones landscape. The papers here by no means do justice to the signifcant body of work in each of these technology domains.

Central Role of Decentralization: The autonomy offered by decentralization is both a boon and a challenge. While it fosters adaptability, it necessitates advanced algorithms for behavior, communication, and optimization.
Navigational and Communication Complexities: Urban environments introduce unique complexities. Efficient algorithms that fuse communication modules with sensors, coupled with advanced obstacle detection methodologies and path planners, can counter these.
Security Imperatives: Commercial drones can be hacked, jammed, spoofed by rogue actors. AES-26 Encryption could be a possibility though this incurs performance tradeoffs.

Conclusion:

Deploying GNSS-independent swarm drones in urban missions is layered with challenges and burgeoning frontiers. From open literature and our recent field experience at the recent Exercise Forging Sabre 2023, we conclude there is an imperative for further exploration. Continued endeavors in developing robust algorithms, enhancing communication protocols, and fortifying security mechanisms will undoubtedly catalyze the full realization of decentralized swarm drones in urban missions.

DSTA Robotics & Autonomy Engineers at Exercise Forging Sabre 2023, at Idaho, USA

References:

Manoni, T., Albani, D., Horyna, J., Petracek, P., Saska, M., & Ferrante, E. (2022). Adaptive arbitration of aerial swarm interactions through a gaussian kernel for coherent group motion. Frontiers in Robotics and Ai, 9. https://doi.org/10.3389/frobt.2022.1006786

Wang, F., Huang, J., Low, K., Nie, Z., & Hu, T. (2022). Agds: adaptive goal-directed strategy for swarm drones flying through unknown environments. Complex & Intelligent Systems, 9(2), 2065–2080. https://doi.org/10.1007/s40747-022-00900-9

Gupta, A., Dorzhieva, E., Ahmed, B., Alper, M., Fedoseev, A., & Tsetserukou, D. (2022). Swarmhawk: self-sustaining multi-agent system for landing on a moving platform through an agent supervision.. https://doi.org/10.1109/icuas54217.2022.9836080

Vaidis, M. and Otis, M. (2021). Swarm robotic interactions in an open and cluttered environment: a survey. Designs, 5(2), 37. https://doi.org/10.3390/designs5020037

Ashush, N., Greenberg, S., Manor, E., & Ben-Shimol, Y. (2023). Unsupervised drones swarm characterization using rf signals analysis and machine learning methods. Sensors, 23(3), 1589. https://doi.org/10.3390/s23031589

Perera, S., Myers, R., Sullivan, K., Byassee, K., Song, H., & Madanayake, A. (2022). Integrating communication and sensor arrays to model and navigate autonomous unmanned aerial systems. Electronics, 11(19), 3023. https://doi.org/10.3390/electronics11193023

Asaamoning, G., Mendes, P., Rosário, D., & Cerqueira, E. (2021). Drone swarms as networked control systems by integration of networking and computing. Sensors, 21(8), 2642. https://doi.org/10.3390/s21082642

Saadaoui, H., Bouanani, F., & Illi, E. (2021). Information sharing based on local pso for uavs cooperative search of moved targets. Ieee Access, 9, 134998–135011. https://doi.org/10.1109/access.2021.3116919

Tong, P., Yang, X., Yang, Y., Liu, W., & Wu, P. (2023). Multi-uav collaborative absolute vision positioning and navigation: a survey and discussion. Drones, 7(4), 261. https://doi.org/10.3390/drones7040261

Bezas, K., Tsoumanis, G., Angelis, C., & Oikonomou, K. (2022). Coverage path planning and point-of-interest detection using autonomous drone swarms. Sensors, 22(19), 7551. https://doi.org/10.3390/s22197551

Wang, Q., Zhuang, D., & Xie, H. (2021). Identification of influential nodes for drone swarm based on graph neural networks. Neural Processing Letters, 53(6), 4073–4096. https://doi.org/10.1007/s11063-021-10583-x

Dimakos, A., Woodhall, D., & Asif, S. (2021). A study on centralised and decentralised swarm robotics architecture for part delivery system. Academic Journal of Engineering Studies, 2(3). https://doi.org/10.31031/aes.2021.02.000540

Papaioannou, S., Kolios, P., & Ellinas, G. (2022). Distributed estimation and control for jamming an aerial target with multiple agents. Ieee Transactions on Mobile Computing, 1–15. https://doi.org/10.1109/tmc.2022.3207589

Trihinas, D., Agathocleous, M., Avogian, K., & Katakis, I. (2021). Flockai: a testing suite for ml-driven drone applications. Future Internet, 13(12), 317. https://doi.org/10.3390/fi13120317

Demir, U. and Ure, N. (2022). A scalable reinforcement learning approach for attack allocation in swarm to swarm engagement problems.. https://doi.org/10.48550/arxiv.2210.08319

Xu, H., Zhang, Y., Zhou, B., Yao, X., Meng, G., & Shen, S. (2021). Omni-swarm: a decentralized omnidirectional visual-inertial-uwb state estimation system for aerial swarms.. https://doi.org/10.48550/arxiv.2103.04131

Ourari, R., Cui, K., Ahmed, E., & Koeppl, H. (2021). Nearest-neighbor-based collision avoidance for quadrotors via reinforcement learning.. https://doi.org/10.48550/arxiv.2104.14912