Machinery Link

Finding New Productivity

(Progressive Farmer image provided by John Deere)

John Deere recently cracked open for a group of visitors the doors to its Intelligent Solutions Group (ISG), in Urbandale, Iowa. Deere's ISG has 2,000 employees worldwide--830 employees in Urbandale--80% of them engineers, and a fair number of them farm. ISG is turning out new software products every 60 days.

We spent time with Mark Chaney, a 23-year Deere veteran who leads product development within ISG's Automation Delivery group. He describes his team's and ISG's mission: reduce complexity, uncover hidden productivity, go faster and maintain precision.


So much of farming is visual, Chaney says. One shade of green tells a story different from another shade of green. But, each plant tells a story. "We like to say make every seed, every drop and every kernel produced count," he says. "Our teams move the focus from having to worry about machine optimization to [optimization of] production systems." Deere is deploying computer vision and machine learning technologies to optimize production. "We are developing new sensing, computing and user experience[s] that enable our customers to reach new levels of productivity," he says, to create an environment for "micromanagement ultimately to the plant level."

P[] D[x] M[x] OOP[F] ADUNIT[] T[]

Two examples of Computer Vision and Machine Learning (CV/ML) already are in the field:

> Active Fill Control for Self Propelled Forage Harvesters automates the loading spout to automatically fill trucks eliminating a tedious operator task.

> Combine Advisor with ActiveVision Cameras uses cameras to feed grain-quality data into an algorithm that maintains loss and grain-quality expectations. It's a capability noticeable as field conditions change. This capability can save $5 to $15 per acre.


"Farmers will be armed with information that they can turn into actionable insights to tend crops at scale like a gardener would tend to each plant in his garden," Chaney says. "Our goal is to reduce this complexity such that it will be [so much] easier to manage and improve crop yields."


Past Issues