Palladyne AI has secured a new U.S. patent aimed at advancing autonomous systems, strengthening its position in AI-driven defense and industrial technologies.
The company announced the issuance of U.S. Patent 12,517,525 B1, titled “Path Creation, Detection and Prediction Using Primitives.” The patent protects Palladyne AI’s proprietary Bayesian Program Learning (BPL) framework, designed to enable intelligent target recognition, autonomous path planning, and real-time behavioral prediction across multi-domain environments, including space, air, land, and maritime operations.
The patented framework focuses on enabling edge-based AI capabilities, allowing systems to operate without reliance on cloud connectivity. This is particularly relevant for defense scenarios where communication networks may be limited or disrupted.
According to Denis Garagic, Chief Technology Officer and co-founder of Palladyne AI, the BPL framework offers differentiated capabilities compared to conventional AI systems. “We believe our patented framework can recognize targets across multiple sensor types without the cloud, convert spoken instructions into optimized robotic motion plans in seconds, and maintain tracking even when signals are lost,” he said.
The patent expands Palladyne AI’s intellectual property portfolio across three key domains. In target recognition, the framework enables detection and tracking of moving objects using multiple sensor inputs, including electro-optical (EO), infrared (IR), LiDAR, radar, acoustic, and radio frequency (RF) systems. These processes are executed entirely on-device, improving response times and operational reliability.
In autonomous path planning, the system can translate natural language commands into optimized motion plans without requiring manual reprogramming. This capability is expected to reduce deployment times in sectors such as manufacturing, logistics, and field operations.
The third area, behavioral prediction, focuses on maintaining track continuity. The system can predict target movement even in conditions involving sensor dropout, signal interference, or occlusion, supporting operations in contested or communications-denied environments.
Ben Wolff, President and Chief Executive Officer of Palladyne AI, described the development as a significant step in the company’s long-term strategy. “This patent represents a fundamentally different approach to machine intelligence,” he said. “We believe it will play a key role as autonomous systems become central to both defense and industrial applications.”
