Mountain View, California, United States
Co-led a team of 20+ engineers across software, ML modeling, data analysis, and hardware orgs to design and deploy a precision agricultural spraying system leading to successful field demo and technology transfer. Peer awards: 25 received, 32 given.
Made core system tradeoffs affecting latency, including camera geometry, sensor type, exposure, optics, flash synchronization, camera parallelism, compute parallelism, sensor readout strategy, and fidelity to training data. The second system improved latency by 3X and pixel throughput by 24X.
Designed and executed numerous lab experiments, including aerodynamic and hydrodynamic performance on a 7 m/s indoor spray track. Hardware in the loop enabled calibration and validation.
Integrated visual, infrared, X-ray, multispectral,thermal cameras, and RTK satellite positioning on a combined fleet of nearly 100 vehicles deployed on three continents.