Syngenta robodog patrols crop rows: quadruped with edge AI spots weeds and disease
Syngenta, in partnership with the University of Nottingham, deploys a quadruped robodog as a field scout: it walks between crop rows without damaging plants or compacting soil (up to 4 hours battery), captures 3D plant imagery with depth cameras, and runs computer-vision models on an onboard GPU at the edge to detect weeds, pests and disease in real time - sending back conclusions rather than raw footage, eliminating the drone-style land-pull-card-upload-wait loop. It fills the crop-level imaging gap left by satellites (field-level), drones (plot-level) and phones (spot checks): satellites see the field, drones the plot, the robodog the leaf. Potential uses include greenhouse monitoring, new-variety yield assessment and agronomist early warning. A representative agriculture case of quadruped robot plus edge AI.
