Skip to main content
ARS Home » Southeast Area » Auburn, Alabama » Soil Dynamics Research » Research » Publications at this Location » Publication #432514
Research Project: Agronomic and Engineering Solutions for Conventional and Organic Conservation Agricultural Systems in the Southeastern U.S.

Location: Soil Dynamics Research

Title: Scalable voice-controlled drone autonomy: From edge-deployed single agents to distributed HITL swarm simulations

Author
item KODE, A - University Of Texas At Tyler
item MANJUNATHA, H - The University Of Texas At Dallas
item BORAH, S - University Of Texas At Tyler
item Torbert Iii, Henry
item SUNDARAVADIVEL, P - University Of Texas At Tyler

Submitted to: Meeting Abstract
Publication Type: Proceedings
Publication Acceptance Date: 3/14/2026
Publication Date: 4/10/2026
Citation: Kode, A., Manjunatha, H., Borah, S.S., Torbert III, H.A., Sundaravadivel, P. 2026. Scalable voice-controlled drone autonomy: From edge-deployed single agents to distributed HITL swarm simulations. Proceedings for 19th IEEE Dallas Circuits and Systems Conference.

Interpretive Summary: This manuscript presents a scalable, voice-driven autonomy framework for unmanned aerial vehicles (UAVs) operating in unstructured agricultural environments, addressing both the cognitive burden of manual telemetry control and the limited scalability of conventional simulation tools on edge hardware. The proposed system is based on a Split-Node architectural topology that decouples high-level semantic intent processing from low-level flight dynamics and state estimation, enabling natural language interaction while preserving deterministic, safety-critical control. The proposed approach enables an “eyes-up” human–swarm interaction paradigm, shifting the operator’s role from low-level piloting to high-level mission command, and provides a practical pathway toward scalable, edge-deployed UAV swarm autonomy for precision agriculture and related applications.

Technical Abstract: This mansucript presents a scalable, voice-driven autonomy framework for unmanned aerial vehicles (UAVs) operating in unstructured agricultural environments, addressing both the cognitive burden of manual telemetry control and the limited scalability of conventional simulation tools on edge hardware. The proposed system is based on a Split-Node architectural topology that decouples high-level semantic intent processing from low-level flight dynamics and state estimation, enabling natural language interaction while preserving deterministic, safety-critical control. In physical deployments, the framework integrates a Raspberry Pi 5 companion computer with a Pixhawk flight controller to support voice-guided autonomous missions, achieving approximately 90% command interpretation accuracy and sub-second end-to-end response latency during controlled field experiments. To extend the system to multi-agent scenarios, a distributed hardware-in-the-loop (HITL) architecture is introduced, in which the Raspberry Pi 5 functions as a headless physics cluster. By incorporating the ArduCopter software-in-the-loop engine, X Virtual Framebuffer (Xvfb) virtualization, and tmuxbased session management, the platform sustains concurrent simulation of eight autonomous agents without graphical overhead. Experimental results demonstrate reliable formation execution, stable telemetry aggregation, and secure low-latency command injection via encrypted channels. The proposed approach enables an “eyes-up” human–swarm interaction paradigm, shifting the operator’s role from low-level piloting to high-level mission command, and provides a practical pathway toward scalable, edge-deployed UAV swarm autonomy for precision agriculture and related applications.