Case Study: Saab Aeronautics achieves real-time drone object detection with Embedl

Saab Aeronautics speeds drone object detection 2.26× with Embedl

Saab Aeronautics, a global leader in defense and security solutions, partnered with Embedl to advance the capabilities of its drone-based object detection systems. The challenge was to deploy a high-performance, real-time deep learning model on a drone's edge device, which presented severe limitations in computational resources, power, and latency, all while preserving critical accuracy.

Embedl leveraged its expertise in model optimization, using techniques like quantization to compress and accelerate the object detection network. This solution delivered a 2.26x speedup in runtime performance while maintaining accuracy with no measurable drop. The improved performance enabled the drone to process real-time video feeds more efficiently, validating the feasibility of advanced AI on constrained hardware and paving the way for future edge AI applications in aerospace and defense.

Saab Aeronautics