Meet MARV, The AI Powered Farmer Robot
summary
A field robot called MARV runs crop trials: it drives at 3 mph spray or slower imaging speed, uses centimeter GNSS for per-plant geotagging, and applies a high‑precision sprayer while logging weather and soil. Vision runs tinyML on Sony sPRiNX modules with Edge Impulse, classifying maize wheat and sugarcane at about 96% accuracy and ~3.8 ms on‑device inference. The design is modular with PoE cameras, battery and wheel prognostics via ML, and optional LoRaWAN LTE 5G or satcom backhaul.
AI-written summary · about
chapters (11)
- 0:00Intro
- 2:05Why is MARV Needed
- 3:50Challenges Several Engineering Challenges are faced
- 6:52The Marv Team
- 7:40MARV's Capabilities
- 10:27Processing On Marv
- 12:03Distributed Processing Approach
- 13:13Distributed Processing Element
- 15:09Distributed Processing Tasks
- 19:15Crop Identification - Live Classification
- 19:59Crop Identification - Hardware Deploy
related talks
27mAI-Driven Smart Farming Optimisation in the Green.Dat.AI Data Space
25mTIS Group’s AI Journey: From Healthcare to Agriculture
25mHow to Accelerate Impact with AI
01h 00mDrones, AI & IOT - What's all the buzz about ?
10mGenesis Final Round - STUDIO LAB
21mIoT Competition Teams
15mLoRaWAN + AI and AI + LoRaWAN From Smart Sensing to Edge Intelligence
27mGenerative AI and IoT, is there a match?